Do all amplifiers sound the same thread

[mtrycrafts]

... I'm just trying to understand where some come from that believe that amps add finer details, tighter bass, broader/wider soundstage ect... I

...

If amps had such effect on those qualities to make them audible, historical DBT data while not peer reviewed would have certainly shown greater numner of amp differences. They do not for a good reason.

 

[mtrycrafts]

From a technical standpoint, this is true. From a real world standpoint, I think the implications are clear: you can't just write off human psychology. Pride of ownership matters quite a lot to most people.

One thing I found interesting was that before the (in)famous Stereo Review test, they let both the "skeptics" and the "believers" spend some time with the amplifiers they were going to blind test; during this session, it was noted that practically all the listeners, including the skeptics who should have (at least in theory) known better, felt they could hear differences.

http://webpages.charter.net/fryguy/Amp_Sound.pdf

Sure, pride of ownership is important or could be the only reason for component choices, no argument there nor requires one. That aspect is not testable or falsifiable. Or, for that matter, a preference or "just because." At least that is what I was trying to get across about not having meaning of sighted comparisons for audible differences. It is about human psychology and how it can get in the way of facts if one is after facts.

 

[mtrycrafts]

....

In a test of listener perceptions of speakers or amps that takes place on a single day, I don't see any advantage that a double blind test has over a single blind test, as long as the tester(s) is being honest.

Well, maybe. If one is not honest, certainly he will convey the answers.
But, that doesn't mean the an honest person will not as it can happen subconsciously that some observers may pick up on.

 

[jinjuku]

could you post the link again. can't find it now

It's in the 'Request for Pro Amp Reviews'. LINKY

 

[mtrycrafts]

...
The next question is what does a 9 out of 10 score mean? 8 out of 10?
I don't think either would be conclusive, but repeating the entire test and scoring 9 out of 10 for the same amp as the first round would make for a significant result.

And, can this box that gene mentioned count the guesses correctly or, of equal importance, randomly swap amps between inputs???

If you don't swap it randomly as in an ABX box, you could keep track of all your switching as there may be only two positions, left and right, either amp A or B. So, at the end, you could just say A 10 out of 10 no matter what, no?

 

[mtrycrafts]

This switcher is made by Audio by Van Alstine avahifi. It isn't listed as an available product on the website, but there are at least a few of those in existence. I have no idea what it might cost. I saw one on loan that Dennis Murphy used last July at the Capital Audio Fest, and it seemed very cool.

Then the question becomes what else can this switcher do besides level match properly and switch silently between A and B? Important aspects may not be part of it.

 

[mtrycrafts]

And that is the reason I couldn't square his statement. He said 'Ever Make'. There is a lot of talent out there. The Crown video I posted they even talk about their current class D's doing stuff better then analog amplifiers on the market. That's 3:57 into the video. That's one bold statement to make and I'm sure the engineering team doesn't make that claim lightly.

I think that interview dates back to 2005, seven years ago, and I'm not sure what class D amps were available then. Perhaps the only thing new with audio amplifiers in recent times is the progress with class D amps.

He has one quote that I can not agree with however:

However, it seems unlikely that one will ever make a Class-D amplifier that can sonically outperform the best-designed Class A-B amps. The newer, less expensive and more compact amplifier technologies may also help foster the migration of the power amplifiers into the cabinets of active loudspeakers.

In my humble and unimportant opinion class D doesn't have to outperform just equal them and you are ahead of the curve, no?

 

[Irvrobinson]

I forgot to mention another silly and probably immaterial test I did. Since the AT602 has the same gain factor as the AT3005 does through the XLR input, I used one channel of the AT602 and one channel of the AT3005 to drive the left and right speakers simultaneously. ATI does a good job on production consistency. Not only couldn't I discern a preference, but on my Stereophile Test CD the pink noise was perfectly centered between the speakers.

The music CD I listened to in this case was this one by Trevor Pinnock. It has violins distinctly and vividly heard on both sides. I know that I couldn't tell which channel was which, or if it was both AT3005 or both AT602.

 

[AcuDefTechGuy]

On the question of double blind vs. single blind test… Although it is true that a tester who is not blinded may unintentionally reveal subtle clues, an honest tester can easily avoid this.

In clinical trials, the real value of double blind studies comes when patients are treated over several weeks or months and may interact with more than one doctor or nurse during their repeated visits. Although medical staff may mean well, expecting them to reveal nothing over a long period of time, and expecting that all such medical staff will operate the same way is asking too much. In such cases, a placebo-controlled double blind study is a more reliable...

Double-blinded clinical trials means none of the doctors or nurses even know which drug is which. Only the pharmacists mixing the drugs know which drug is which. And we all know that pharmacists are too honest to reveal anything. So the physicians and nurses would not be able to reveal anything to the patients even if they wanted to.

And placebo trials are considered unethical.

However, single-blinded amp/ speaker studies are better than nothing. We'll take them.

 

[cpp]

As far as blind testing, the people being tested should not know what equipment is being tested and the tester should not acknowledge a device type. In all testing its still up to the people doing the test to ensure results are honest and not shaped to support a given product. And then what does this really prove as nobody's hearing is the same, nor is the appreciation for the music being used.

 

[lavath]

I have noticed differences in how I perceive the sound from digital amps compared to tube amps. not sure that analytical gear can capture the ear differences...

sorry but I need 5 posts to show my home theater build page I made

 

[jinjuku]

I have noticed differences in how I perceive the sound from digital amps compared to tube amps. not sure that analytical gear can capture the ear differences...

sorry but I need 5 posts to show my home theater build page I made

Tubes behave in a different and predictable manner vs SS. They are more graceful going into clipping and their harmonics are often odd order.

 

[moves]

I think the point of contention over blind testing is a simple matter of semantics.

In the literal sense blind means without sight.

However, when doing research involving people, a blind test is one where the subject has no way to know which condition they are experiencing. By this definition and with Gene's statement above, this would be a blind test.

A DBT (double blind test) requires that the person administering the test also be "blind" to which condition is which. This is the standard for good testing of subjective perceptions. It has been established that a blind test and a double blind test often have different results; while DBT's are repeatable. If the person who administers the test knows which condition is which he will unintentionally give off subtle cues which are subconsciously picked up by the subjects in the test.

In the case of testing one amp against another, there is no reason for the test administrator to be in the room with the subject (assuming the subject can be trusted to not investigate the wiring), so it would be easier to do an "Indirect DBT". I would propose the test administrator wire up an A/B switch and leave the room. The subject then would enter the room and establish which amp, A or B, was preferred and write it on a pad. The subject would then leave the room before the administrator saw the result. Some type if bell or simply a single knock on the wall could communicate when one party had completed their task. Repeat this 10 times. If the score was then 10 to 0 for one amp vs the other, it is done. Anything else, repeat the ten trials a second time.

Understand that I am no expert at this type of testing. There is probably a better protocol. I wanted to get specific to make the concepts clear and to throw out a discussion document for refinement (if that is a place this thread wants to go).

In case someone doesn't fully understand the need for the double blind test, simply consider that if the test administrator receives the info from the third trial and it is consistent with the result from the first two (a score of 3 to 0). He might (without awareness) raise an eyebrow or betray some subtle hint of congratulation or respect to the subject.

The next question is what does a 9 out of 10 score mean? 8 out of 10?
I don't think either would be conclusive, but repeating the entire test and scoring 9 out of 10 for the same amp as the first round would make for a significant result.

Thinking scientifically, I think that any score other than 10/10 would introduce some skepticism about the question at hand.

 

[moves]

I like point made about how amps depend on the load of the speaker as well as the size of the room it is playing in. I agree that it would almost be impossible to hear a difference between a high end amp and a mid or even low end amp and low listening levels.

 

[PENG]

Tubes behave in a different and predictable manner vs SS. They are more graceful going into clipping and their harmonics are often odd order.

I thought more people may find even harmonics more pleasing. There are probably more tube amps that generate more even harmonics than solid state amps. It depends on the topology of the design that affects the spectrum of the harmonic distortions.

 

[PENG]

Assuming a perfectly resistive speaker load, low noise, low output impedance and the amplifier can maintain sufficient output voltage without power supply lag, then its very conceivable the sonic differences between two amps will be indistinguishable. Lots and LOTS of caveats for this to be the case.

Well, it depends also on the individual's need in the real world. For example, I listened to around 70 to 80 dB level (average/typical) at the most and in my room that typically means my amps output would cruise at around 0.1 to 0.2W (verifiable and confirmed, by me). I will reconfirm the numbers later.

For Diana Krall type of music the DR is not much more than 12 dB or so. Whether I use my 3805 or separates to drive even my 86 dB/W/M with impedance far from being a perfect resisistive load, there is no audible difference, blind test or not. I do believe the power vs sound quality thing is real, but often get exagerrated to the nth degree, by that I mean for most real world home applications, barring those in a large room, sit 4 meters or more from the speakers and enjoys 85dB or higher average SPL.

 

[Swerd]

Double-blinded clinical trials means none of the doctors or nurses even know which drug is which. Only the pharmacists mixing the drugs know which drug is which. And we all know that pharmacists are too honest to reveal anything. So the physicians and nurses would not be able to reveal anything to the patients even if they wanted to.

In a properly conducted DBT, the patients receiving the drugs never see the pharmacists who know the placebo from the experimental drug. And those pharmacists never know who is getting what. The bottles only have patient ID numbers on them. Medical staff, who are blinded, receive the drugs from the pharmacists and give them to the patients. Until the trial is unblinded, no one else may know who gets what.

And placebo trials are considered unethical.

Not true. If there is a standard treatment for a disease, any clinical trial of patients with that disease must also include the standard therapy. For example, if a trial tests drug X in patients with lung cancer, and the standard first-line therapy for lung cancer is carboplatin plus paclitaxel, a trial would treat lung cancer patients with drug X only after they had first gotten the standard therapy and their disease progressed.

If that showed positive results, a more elaborate and expensive randomized placebo-controlled trail would be done. It would have two arms. Arm A could be the standard therapy plus drug X, and arm B could be the standard therapy plus a placebo.

Anything else would be unethical, and would never be approved by a hospital review committee or the FDA. Any MD who did such an unethical trial despite those safeguards in the cancer field, would loose his accreditation and probably could never work as an oncologist again. Such things have happened.

However, single-blinded amp/ speaker studies are better than nothing. We'll take them.

Yes . And that was my point.

 

[JerryLove]

I have noticed differences in how I perceive the sound from digital amps compared to tube amps.

It's possible. Tubes clip more gracefully than most solid state amps. Because tube amplifiers tend to be capable of less power, they will be clipping more often into a given loudspeaker.

not sure that analytical gear can capture the ear differences...

Mics can hear underground nuclear testing from the other side of the world. They can form 3d pictures of the surrounding area from sound. They can listen to conversations from hundreds of years away.

21st century technology is far *more* capable of detecting variation in waves than your ears are; not less.

 

[Irvrobinson]

Tubes clip more gracefully than most solid state amps.

Lots of solid state amps have so-called "soft-clipping" circuitry that keeps them from ever actually clipping, making their performance better in this regard than tubes.

 

[JerryLove]

Lots of solid state amps have so-called "soft-clipping" circuitry that keeps them from ever actually clipping, making their performance better in this regard than tubes.

More accurate. More true-to-source. "Better" is subjective. Some people *like* the distortion/coloration that tube amps provide.