Can you hear a difference in Sound between Audio Amplifiers?

[Irvrobinson]

Long live blind testing.

A guy walks up to a man dressed up in a weird costume, who is dancing around and jumping around, in some sort of dance. He's dancing to bongo drums played by a young boy. The guy stands there and watches the dance until the man in the costume stopped.

"Why are you dancing?"

"We're in a drought. My well went dry. It's a rain dance."

"Have your rain dances worked in the past?"

"No."

"Then why are you dancing?"

"The drought is pretty bad, and this is only thing I can think of to do."

 

[Swerd]

A guy walks up to a man dressed up in a weird costume, who is dancing around and jumping around, in some sort of dance. He's dancing to bongo drums played by a young boy. The guy stands there and watches the dance until the man in the costume stopped.

"Why are you dancing?"

"We're in a drought. My well went dry. It's a rain dance."

"Have your rain dances worked in the past?"

"No."

"Then why are you dancing?"

"The drought is pretty bad, and this is only thing I can think of to do."

You can do better than that. The rain dance story is about making false correlations (except the dancing guy admits it hasn't worked yet).

Blind tests of two different brands of aspirin could have confusing results because both are essentially the same. But if aspirin was tested blind against a placebo, among people with a fever, the results would show a clear difference. It depends on how you design the test.

I realize I won't convince you about this . I feel a little like the guy who asked the rain dancer why he does it.

 

[Irvrobinson]

You can do better than that. The rain dance story is about making false correlations (except the dancing guy admits it hasn't worked yet).

Blind tests of two different brands of aspirin could have confusing results because both are essentially the same. But if aspirin was tested blind against a placebo, among people with a fever, the results would show a clear difference. It depends on how you design the test.

I realize I won't convince you about this . I feel a little like the guy who asked the rain dancer why he does it.

No, Swerd, as a member of the blind testing cult for audio electronics, you are the rain dancer. I think audio electronics do sound essentially alike, I just don't think blind testing adds any value when humans have such poor audio memory, and the input to the process are mere guesses. Fever reduction and reduced pain are not guesses (though the latter is subjective).

 

[auronihilist99]

Hello Irvrobinson,

I don't understand your rain dance analogy when it comes to volume level-matched double-blind listening tests with audio components.

 

[Swerd]

No, Swerd, as a member of the blind testing cult for audio electronics, you are the rain dancer. I think audio electronics do sound essentially alike, I just don't think blind testing adds any value when humans have such poor audio memory, and the input to the process are mere guesses. Fever reduction and reduced pain are not guesses (though the latter is subjective).

My major point has been that electrical lab bench measurements of electronic audio gear (no matter how well done), can never directly answer the question of what actually is and isn't audible to humans. Lab bench tests may predict the answer, but only listening tests can directly answer that question.

Perhaps you have never seen a properly designed blind test. If all the necessary positive and negative controls are built in, useful conclusions are easier to make. Especially if the conclusion is "No we cannot hear a difference". However, I make no promises to ever conduct such a trial.

I would never advocate widespread blind listening tests for audio electronics because it would be like listening among different brands of grass seeds to see which one you could hear growing.

I'd make a very poor living as a rain dancer as I only bet on sure things.

I am not a cult member. I do strongly advocate speakers that have flat frequency response curves, but I'm happy with nearly any amp as long as it has enough juice to drive speakers without clipping.

 

[Swerd]

Hello Irvrobinson,

I don't understand your rain dance analogy when it comes to volume level-matched double-blind listening tests with audio components.

We're engaging in friendly chop-busting over fine points. I'm advocating only truth and science while I strongly deny being a rain dancer . And Irvrobinson is being a curmudgeon .

Personally, I'd accept accusation of being a curmudgeon sooner than being accused of being a rain dancer .

 

[RichB]

What I found interesting about the Parasounds is that there was a lot of protection circuitry, which according to audiophiles is a no no, there was, I believe no fewer than 5 fuses in the amp.

A fair point. Even John Curl has stated on message boards that he prefers breakers, while seeming to give a nod to silver fuses. A fuse is resistive and if you want to focus on the power source, it seems a more likely candidate than a power cord. I ran 20 amp circuits and use stock cords.

I blew a couple of fuses on the Parasounds, for no apparent reason. Just turned them on.
This is part of the reason that I moved to ATI which uses circuit breakers. I found only one fuse on the inside the AT6000 which appears to protect the logic board DC supply. There may be others hidden under the heat-sinks, I cannot be sure.

I have had the following amps driving my center and surrounds: Sunfire Signature Grand Signature, Sunfire 7400, ATI3000, Parasound A51, and now ATI Signature ATI6000.

The ATI 6000 amps is the first amp that has no audible hiss or hum on the center or surrounds. The source of the problem is likely environmental, but the ATI 6000 finally cured it.

I have never tried Pro amps but they would not meet the aesthetic criteria of the "other" decision maker. The AT6000's barely squeaked by.

- Rich

 

[auronihilist99]

That ATI AT6002 is a beautiful amp. It looks very well made too. I can't help but wonder what the outcome would be when performing a volume level-matched double-blind test between the ATI and a Denon or Yamaha AV receiver?

 

[Irvrobinson]

That ATI AT6002 is a beautiful amp. It looks very well made too. I can't help but wonder what the outcome would be when performing a volume level-matched double-blind test between the ATI and a Denon or Yamaha AV receiver?

Can't say, but I did a comparison between my formerly top of the line, fully-balanced ATI AT3005 in stereo mode (meaning, using only two of the five channels), and my bottom of the line, single-ended ATI AT602 driving a pair Revel Salon2s (same as RichB's) full range. I have a pretty large room, and playing at realistic volume a live recording of my wife's largest drum kit (engineered by me in a former house), and the AT602 easily clipped, while the AT3005 played the recording significantly louder, which was much more realistic. (A drum kit in a home is loud.) Other than that, I could discern no differences whatsoever. Nada. Same silent background noise level at idle. Same great sound. Within the AT602's power envelop, same great bass.

 

[PENG]

That ATI AT6002 is a beautiful amp. It looks very well made too. I can't help but wonder what the outcome would be when performing a volume level-matched double-blind test between the ATI and a Denon or Yamaha AV receiver?

I have compared my A21 and Bryston 4BSST with my Denon 3805 and found no easily audible difference at volume that I can withstand, not in any blind test for sure. The A21 is comparable to the ATI in terms of specs and price. The Denon actually seemed to be the quiestest. My guess is that in a blind test the Denon 4000, Yamaha RX-A2000, or Marantz 6000 series and above will hold its own against the ATI. I would bet that's not the case for the entry level Denon or Yamaha AVRs.

 

[herbu]

I just don't think blind testing adds any value when humans have such poor audio memory

I've never heard a well executed DBT, but I suppose the ability to immediately switch between systems during playback. I imagine sitting with a single A/B switch in front of me so I can compare a single passage, back and forth, w/o needing any particular "audio memory". Am I wrong?

 

[Goliath]

No, Swerd, as a member of the blind testing cult for audio electronics, you are the rain dancer. I think audio electronics do sound essentially alike, I just don't think blind testing adds any value when humans have such poor audio memory, and the input to the process are mere guesses. Fever reduction and reduced pain are not guesses (though the latter is subjective).

That's why quick-switching exists and is often implemented in sensory testing.

If blind testing has no value in your eyes then I shudder to think what your opinion of a totally uncontrolled, perceptually error-prone "seeing/knowing/expecting.... bias-ridden etc" evaluation is.

Even if blind tests were fatally flawed, what does that make a casual sighted evaluation?

 

[Irvrobinson]

I've never heard a well executed DBT, but I suppose the ability to immediately switch between systems during playback. I imagine sitting with a single A/B switch in front of me so I can compare a single passage, back and forth, w/o needing any particular "audio memory". Am I wrong?

Actually, what you need is an A/B/X switch, so that sometimes you test the difference between A and A. However, even if you just did A/B switching you would find that audio memory is a factor, unless you're switching in a matter of seconds. Are you also assuming something like restarting an audio track over and over after, say, 30 seconds of playing time? Just saying, this sort of testing can get pretty complicated fast. Finally, comparative testing can test for preference, but it doesn't test for accuracy. Accuracy tests require a reference.

 

[Irvrobinson]

That's why quick-switching exists and is often implemented in sensory testing.

If blind testing has no value in your eyes then I shudder to think what your opinion of a totally uncontrolled, perceptually error-prone "seeing/knowing/expecting.... bias-ridden etc" evaluation is.

Even if blind tests were fatally flawed, what does that make a casual sighted evaluation?

Ah Goliath, back for another round of drive-by obnoxious argument? If you bothered to read the thread you would see that I am not arguing for casual sighted evaluation.

 

[AcuDefTechGuy]

I am not arguing for casual sighted evaluation.

Let me see if I get your point.

1. Medical/drug double-blind testing is absolutely important in life.

2. Double-blind testing in the audio/video hobby is overkill and not really important in life. We can live without it. Some people just won't believe the results anyway no matter what. And because it is not regulated (like medical drug testing by FDA), the flaws may be significant.

 

[Goliath]

Ah Goliath, back for another round of drive-by obnoxious argument? If you bothered to read the thread you would see that I am not arguing for casual sighted evaluation.

Yet another affirmation of what Dunning-Kruger found...

If you are going to argue against blind testing then please make an attempt to come up with (or imagine) a reasonable argument. Memory is not a valid objection. Enter quick-switching.

Next excuse/red herring for why blind testing is ineffective in 3...2...1....

 

[Irvrobinson]

Let me see if I get your point.

1. Medical/drug double-blind testing is absolutely important in life.

2. Double-blind testing in the audio/video hobby is overkill and not really important in life. We can live without it. Some people just won't believe the results anyway no matter what. And because it is not regulated (like medical drug testing by FDA), the flaws may be significant.

I agree with both of your points, but my point about audio *electronics* blind testing is that comparative testing, blind or sighted, is flawed because humans don't have sufficiently accurate hearing acuity or audio memory to discern between properly designed and functioning electronics being compared, so the responses to the test are generally guesses, which are garbage responses. Put another way, I think most electronics do not have an audible signature, but blind testing is not a good way to prove it.

For speakers, which generally do have audible differences, I think blind testing can be helpful in detecting preferences, but not accuracy.

 

[AcuDefTechGuy]

I agree with both of your points, but my point about audio *electronics* blind testing is that comparative testing, blind or sighted, is flawed because humans don't have sufficiently accurate hearing acuity or audio memory to discern between properly designed and functioning electronics being compared, so the responses to the test are generally guesses, which are garbage responses. Put another way, I think most electronics do not have an audible signature, but blind testing is not a good way to prove it.

For speakers, which generally do have audible differences, I think blind testing can be helpful in detecting preferences, but not accuracy.

I agree that testing in this hobby (speakers and electronics, blind or not) will never be close to being as valid as large nationwide randomized medical double-blind testings.

But I guess the question is, in this hobby, electronics or speakers, even when it is flawed, is blind testing still better than sighted testing?

I will say this. I don't like it when the subjects in the tests (blind or not) are trained to hear certain things or taught "how to listen critically".

I just want to know what people prefer. They don't need to be told what to listen for and what they should prefer. That's inducing bias IMO.

And if we are talking about preferring the sound of one speaker over another, I'm not sure how valid the results would be. I mean at the end of the day, people just care about what sounds best to them, right? Do they care if it sounds good to 20 other guys who prefer a certain speaker?

 

[jinjuku]

No, Swerd, as a member of the blind testing cult for audio electronics, you are the rain dancer. I think audio electronics do sound essentially alike, I just don't think blind testing adds any value when humans have such poor audio memory, and the input to the process are mere guesses. Fever reduction and reduced pain are not guesses (though the latter is subjective).

Irv you can't have it both ways. Poor audible memory that negates bias controlled testing also negates sighted evaluation. Period.

So which one is it?

 

[Irvrobinson]

Irv you can't have it both ways. Poor audible memory that negates bias controlled testing also negates sighted evaluation. Period.

So which one is it?

You too? I am not advocating sighted comparisons either. I think they're both ineffective. (With the exception of the idle hiss that both you and I have experienced with some products.) I think we're stuck with measurements.