Can you hear a difference in Sound between Audio Amplifiers?

[AcuDefTechGuy]

And, you pass on the coded drugs that only you know what they are, placebo or the real stuff, to the administrator, than he in turn to the recipients in the study. At the end of th estudy, notes are compared which drug wuoked how effectively, etc.

1. We don't use placebos anymore since it is considered unethical.
2. The pharmacist hands the drugs to the pharmacy technician. The technician hands the drugs to the nurse. The nurse administers the drugs. The nurse aid measures the vital signs (blood pressure, etc) or collects blood. The lab measures the blood, etc.
3. And there are hundreds of subjects in several hospitals.

Yeah, drug DBTs are way more extensive than audio.

 

[mtrycrafts]

Absolutely. What's the point, if at first you don't test the testers?

Oh, but they are the ones usually claiming to hear difference. Do you want to disqualify them due to poor hearing and burst their bubble?

 

[mtrycrafts]

I can see it now, a blind testees test. Would that be single or double blind?

That depends on how many eyes are closed.

 

[mtrycrafts]

I think, or at least I use the term to imply a person's bias, not statistics or implementation bias, and perhaps the subconscious bias of wanting it to happen to your liking.

 

[Boerd]

For short-term comparisons I lean your way, but simple specs like you're listing don't tell the whole story. The frequency distribution of noise and distortion orders make blanket statements tougher to defend. But again, I wouldn't bet a nickel on my ability to differentiate between excellent amps on demand. Aural memory just stinks. If only we could compare two audio presentations side-by-side the way we can with video.

Hmm - no we can't do such a comparison with audio. But blind A-B should be close enough.
Now, even blind A to B comparisons will not be able to make the case for 50k$ amps. These days 1k$ buys you an amp that's impossible to discern in normal listening conditions from the uber-priced category.

 

[Alex2507]

These days 1k$ buys you an amp that's impossible in normal listening conditions to discern from the uber-priced category.

Now that is something I would like to see put to the test.

 

[mtrycrafts]

For short-term comparisons I lean your way, but simple specs like you're listing don't tell the whole story. The frequency distribution of noise and distortion orders make blanket statements tougher to defend. But again, I wouldn't bet a nickel on my ability to differentiate between excellent amps on demand. Aural memory just stinks. If only we could compare two audio presentations side-by-side the way we can with video.

Even with video you still need some memory unless you pause the image that you just cannot do with audio unless you use a test tone that doesn't change. And then, you have to scan the image back and forth comparing an area you think is different, no?
But, come to think of it, how would your brain interpret two simultaneous inputs to the brain at the same time, a stereo image that may not have any comparative meaning? Interesting dilemma.

 

[Irvrobinson]

Even with video you still need some memory unless you pause the image that you just cannot do with audio unless you use a test tone that doesn't change. And then, you have to scan the image back and forth comparing an area you think is different, no?

No. Audio and video are completely different. Freeze video and you get a photograph, which can be used for detailed comparison. And your eye moving from one image to another is quicker and more effective than any audio A/B comparison.

 

[PENG]

No. Audio and video are completely different. Freeze video and you get a photograph, which can be used for detailed comparison. And your eye moving from one image to another is quicker and more effective than any audio A/B comparison.

Agree, just a side note, that's actually why I think the claim that there are things in audio amplifiers that cannot be measured as the reason why audible differences not shown by published and measured specs is not valid, because any such differences could have been heard by instruments, then seen and proved to be audible by humans or not. It is proved by advanced mathematical analysis on the recorded waveforms/data so even if human eyes cannot see the differences, they can see the analytical results that could be visible. Put it that way, if the best recording instruments and spectrum analyzers currently available cannot record and show it, we cannot hear it, unless we don't believe in landing on the moon or some of us are X-men/Super man.

 

[cpp]

just a side note, that's actually why I think the claim that there are things in audio amplifiers that cannot be measured as the reason why audible differences not shown by published and measured specs is not valid, because any such differences could have been heard by instruments, then seen and proved to be audible by humans or not. It is proved by advanced mathematical analysis on the recorded waveforms/data so even if human eyes cannot see the differences, they can see the analytical results that could be visible. Put it that way, if the best recording instruments and spectrum analyzers currently available cannot record and show it, we cannot hear it, unless we don't believe in landing on the moon or some of us are X-men/Super man.

Measurements taken of audio devices are just numbers that do not represent what a human could or might hear. A dog might hear it or a elephant might hear it but some of these measurements are so far above or below what a normal non-superhero can hear. So are these woopie-doo SUPERHERO only measurements just meaningless attempts at good PR for sales ? and what does this have to do with blind testing amps and who has all of these high end scopes and audio measuring devices to measure with anyway and are they the best of the best because we sure don't want to use "just your everyday device", because someone might comment that the measuring equipment was sub par. now for a cup of coffee

 

[3db]

The whole sighted test bias argument is a bit overstated in my experience. I've seen plenty of very biased "blind" tests that were a lot less valid then well controlled "sighted" tests. Anytime you precondition a listener by either bragging of a brand, price or what they'd expect to hear will undoubtedly bias their perceptions of the products. However, if the listener doesn't know brand, or price and you don't tell them which amp is playing, then there is NO reason to do the test blind.

I'm trying to understand your explanation here. If the listeners are exposed to predetermined expectations of one amp over another but do not know when its being played or not, where does the bais creep in?

I've found when I bring in unexperienced listeners that don't know brands, they don't even care which speaker they are listening too. I can't tell you how many times I've had listeners prefer the uglier, less prestigious brand in a sighted test. Don't let all of the DBT jargon some manufacturers like to throw around "Bias" you against controlled sighted tests. Most companies that claim to run DBT's never actually do. At best they run a SBT and they usually use their own listening panel and never disclose the resultant familiarity bias in those cases.

Speaker preference is subjective..... Shared with room acouticss, its the other 50% of sound a listener will hear. I personally would pick the ugly speakers too if they sounded better.

 

[Irvrobinson]

Agree, just a side note, that's actually why I think the claim that there are things in audio amplifiers that cannot be measured as the reason why audible differences not shown by published and measured specs is not valid, because any such differences could have been heard by instruments, then seen and proved to be audible by humans or not. It is proved by advanced mathematical analysis on the recorded waveforms/data so even if human eyes cannot see the differences, they can see the analytical results that could be visible. Put it that way, if the best recording instruments and spectrum analyzers currently available cannot record and show it, we cannot hear it, unless we don't believe in landing on the moon or some of us are X-men/Super man.

In principle I couldn't agree more. The question is whether or not we have the right tests yet. For example, looking at plots of intermodulation distortion tests... they sometimes reveal very interesting behavior. I look at a typical THD + noise curve and I think, how could that possibly be audible? I look at an FFT at 1W for some amps and wonder if there are artifacts that could be audible, or at least have the fatigue-causing impact that I muse about. And, the problem isn't really testing amps, the problem is with testing humans. I'm just not a fan of DBTs for audio, yet I'm at a loss, obviously, to recommend a different method. My interest, personally, is to just keep looking at different kinds of measurements, and hope that will point us to constructing a test that will make audibility more decisive.

 

[mtrycrafts]

1. We don't use placebos anymore since it is considered unethical.
2. The pharmacist hands the drugs to the pharmacy technician. The technician hands the drugs to the nurse. The nurse administers the drugs. The nurse aid measures the vital signs (blood pressure, etc) or collects blood. The lab measures the blood, etc.
3. And there are hundreds of subjects in several hospitals.

Yeah, drug DBTs are way more extensive than audio.

I can see it being unethical if you are researching an illness that has death as a consequence of no treatment. But, for instance something with no such consequences, how do you know if the drug works? Your way measures it effects against another drug. Either it is less effective and cost more, equivalents or more effective. I guess if it is less effective it is like comparing it with a placebo?

 

[mtrycrafts]

Eureka!

I am reading an article that has an interesting parallel to our discussions here that should be reviewed.
An interesting read from history on the subject of "bias" interaction between experimenter and the subject, be it a human or in the case of the article, an animal. Clever Hans - Wikipedia, the free encyclopedia

Back in the early 1900 Oskar Pfungst experimented about this, "with knowledge" and "without knowledge."
And, here we are a 100 years later and still trying to convince people on the vagaries of experimental weaknesses.

 

[AcuDefTechGuy]

I can see it being unethical if you are researching an illness that has death as a consequence of no treatment. But, for instance something with no such consequences, how do you know if the drug works? Your way measures it effects against another drug. Either it is less effective and cost more, equivalents or more effective. I guess if it is less effective it is like comparing it with a placebo?

If your new drug does not even work at least as well as the current top drugs in the market, you have lost the game.

For example, if your new drug cannot lower blood pressure or blood cholesterol as much as the top current meds in the same class and be as safe or safer with less adverse effects, why would anyone want to take your drug?

Just like in audio. We want to compare to the gold standards of Revel, KEF, TAD, etc, not to Bose.

 

[mtrycrafts]

If your new drug does not even work at least as well as the current top drugs in the market, you have lost the game.

For example, if your new drug cannot lower blood pressure or blood cholesterol as much as the top current meds in the same class and be as safe or safer with less adverse effects, why would anyone want to take your drug?

Just like in audio. We want to compare to the gold standards of Revel, KEF, TAD, etc, not to Bose.

Well, yes we do, but $$$ can be an issue. And, in medicine as well. There are more than one drugs to treat an ailment. Generics do as well and cost way less that the gold standard meds.

 

[LAB3]

I've answered the poll based on the premise that the load applied to the different amps are the same and well within the designed parameters of the "weakest's amp, that levels are matched and that output levels are also well within the weakest amp's power delivery. Yes they will sound the same. However, if one of those variables change, then I expect to hear differences.

I agree. I have owned my speakers for 30 odd years and had tubes and a ss and played them both at 80DB before I sold the tube amp. They both sounded great on my Cornwalls(they don't require hugh amps). My tube friends were Suprised when they came over and listened to my Yamaha HD AVR with their whimpy amps (one of them said before the test). I have a very small den, speakers are 7' apart and towed in some and sofa 7' back.

 

[Hyrlyfrm]

Here's an idea. We have the ability to record a song played on a specific set of gear. That recording can be analyzed as a waveform. After testing amp gain to insure level matching, record a specific song on amp A. Now switch nothing in the system but the amplifier to amp B. Record the same track at the same level as was played using amp A. Now compare the waveforms from amp A and amp B. If there are differences, the waveforms should vary. This idea is of course assuming that there is some audio software out there that can simply overlay and compare the waveforms.

I'd love to know if this is actually possible or if it's just a dumb idea.

 

[PENG]

In principle I couldn't agree more. The question is whether or not we have the right tests yet.

The only tests measurements we can easily find tend to be those published by Stereophile. I do think high end manufacturers such as McIntosh, Krell or even Bryston would have done all the tests and measurements they need in order to convince themselves they have reached the practical point where further improvement in audio specs and performance would not be discernible by humans, but I have no proof.

For example, looking at plots of intermodulation distortion tests... they sometimes reveal very interesting behavior. I look at a typical THD + noise curve and I think, how could that possibly be audible? I look at an FFT at 1W for some amps and wonder if there are artifacts that could be audible, or at least have the fatigue-causing impact that I muse about. And, the problem isn't really testing amps, the problem is with testing humans. I'm just not a fan of DBTs for audio, yet I'm at a loss, obviously, to recommend a different method. My interest, personally, is to just keep looking at different kinds of measurements, and hope that will point us to constructing a test that will make audibility more decisive.

I think one of the problem is that audio amp circuits/designs are really not considered to be something worthy of serious research by universities all over the world. They have better things to do and research in finding out audio amp facts that interested you and I will not get them any grant. I bet you most EE students had to cover audio amps in their electronic courses but again it is never considered as anything that required education level higher than college or first/second year degree levels at the very most. When I was in college (prior to U), audio amp circuit designs were in fact covered and we did build one SS amp as part of the lab work. If I remember right, that was only covered in the first year, after that of course we moved on to "higher" level stuff.

I would love to see any university research papers on the subject are talking about. As for all these talks about ST, SBT, DBT, it seems silly if it is such a big deal to prove a point yet it somehow manage to attact so much attention. One would think that people should just focus more on what counts most, i.e. speakers, source material, room acoustics etc. that have far more impact on audible differences.

 

[Swerd]

I would love to see any university research papers on the subject are talking about. As for all these talks about ST, SBT, DBT, it seems silly if it is such a big deal to prove a point yet it somehow manage to attact so much attention. One would think that people should just focus more on what counts most, i.e. speakers, source material, room acoustics etc. that have far more impact on audible differences.

Exactly.

I've gotten bored thinking about how to show that any differences among amps are essentially inaudible. It's like comparing brands of paint by listening to them dry.

Years ago I used to hang out at another forum, where this genuine lunatic (a moderator at a place called "Cable Asylum") would occasionally show up just to pick fights with everyone. He wasn't really a troll because he identified himself and really believed in all his nonsense. It could get old quickly, but in small doses, it was a lot of fun .