Do all audio amplifiers really sound the same???

Status
Not open for further replies.
Jump to Latest
#821
Pan said:


I think maybe you misunderstod something here.. maybe I was not clear enough. First, it was Jan who mentioned that he use to feel some anxiety. Second I said I had no such problems in connection to listening tests. When our discussion then continued I wrote some about how I function in life in general.. so if you reread the latest couple of posts maybe it become clear.

And hey, next time help Jan out feeling more relaxed and comfortable okay? 🙂


/Peter
Click to expand...


Yes, I brought it up, as the reason that is almost always given by people who are against dbt: "The reason that many dbt come up with a negative result (no differences heard) is that the 'stress' of having to come up with 'a' answer kills their critical listening abilities".

Maybe the word 'anxiety' is too strong, what I mean was some type of stress that lets you focus better and try better. SY, when I say you do that test, you were quite intent, eyes closed, focussed. That *is* a certain stress, but in my book it has a positive effect, like the stress you have if you have to do an important presentation to a bunch of people.

At any rate, the important thing I picked up from this discussion that there are apparently many people that can do either a sighted or a db test in a relaxed way, so the argument: "the stress kills my critical listening abilities" looks like a dodge.

Jan Didden
 
#822
Pan said:
[snip]It's been done many times allready. I have only participated in BT on CD players and microphone preamps though.

Again, the "Swedish audio society" (LTS) has done many blind tests on amps and other stuff. Only a small number (three I think) of many amps that have been tested have prooven transparent in the test set up. That should be read that most amps on the market color the signal audibly.

If I remember correct only one have prooven to be fully transparent on all levels and that is one of the top amps from Bryston. NAD 208 and one of the Halcro amps I believe was mostly transparent but could be pressed to show some small colorations.

The Bryston was not transparent from the factory so the guys searched for the reason for this coloration, reported back to Bryston. Bryston performed some listeing tests and came to the same conclusion and made a change to the amp. An amp was sent to Sweden for further test and prooved to be the best amp ever tested during several decades. No, I'm not affiliated with Bryston.

The findings mirrors (to a high degree) those that was presented in that article by Paul Frindle earlier in this thread.

I also think that the engineers performing these tests always find a correlation between an audible coloration and bench performance. The key is to know how to measure and what to measure.


/Peter
Click to expand...


I have been trying to find that Paul Frindle article - can't find it. Any pointer?

That Swedish article - I guesss that's in Swedish...?

Jan Didden
 
#825
eStatic said:


That being the case and the fact that we can all hear differences in speakers and that measurements unequivocally show speakers to be, by far, the weak link in the audio chain, it seems inescapable that one should first go for the best speaker one can build or buy and spend a relative pittance on the amp until such time as one has more dollars than sense. Then one could decide if the cost of marginal/imaginary improvement was more important than the good that money/time-effort could do as a donation to a worthy cause--like diyAudio.com 😉
Click to expand...

I agree well said. I can't imagine people who love Magnepan, Apogee, or Kef all bringing the same issues to the table on amplification.

A signal processing expert from Bell Labs pointed out to me that two speakers and a classic stereo source can be proven mathematically to be incapable of reproducing exactly a given 3D sound field. I don't remember the reference, but in my mind that always puts in an element of illusion.

My take on amps is low noise/artifact(EMI/RFI) and effortless dynamic range/headroom for a given speaker. The later has been violated in virtually every 'listening' test I ever participated in (sometimes absurdly).
 
#826
scott,

A signal processing expert from Bell Labs pointed out to me that two speakers and a classic stereo source can be proven mathematically to be incapable of reproducing exactly a given 3D sound field. I don't remember the reference, but in my mind that always puts in an element of illusion.
Click to expand...

Well theres hardly a need for such an expert to point out something as obvious. However the illusion of a 3D space between and behind the speaker set up can be frightingly good. For a full soundfield to be reproduced you need many channels. I'm about to start out with eight channel recordings to see what it gives.

My take on amps is low noise/artifact(EMI/RFI) and effortless dynamic range/headroom for a given speaker. The later has been violated in virtually every 'listening' test I ever participated in (sometimes absurdly).
Click to expand...

Yup, low noise is important it seems and I assume you are thinking on dynamic noise primarly and not the static background noise.


/Peter
 
#827
-originally written by janneman
<snip> ...but isn't the 'stress' at non-blind, sighted tests similar? You still have peer pressure (assuming you do this in a group) of not making a fool of yourself and making the wrong decisions. I'don't see that that is radically different between sighted and dbt testing. And if the test is done individually, anonimously, it is completely removed.

Then there is the stress of being forced to make a decision one way or another. Again, isn't that the same when doing sighted or dbt's? At any rate, does that deminish the ability to hear with sensitivity and accuracy? <snip>
Click to expand...

Just to address the first point; group dynamics are factor the experimentator should be aware of and of course i wouldn´t trust a "newbie" in a sighted test.

I wouldn´t call it "stress" just beeing unfamiliar with the blind test situation.
Afaik that is a point most of the people who have conducted DBTs agree in; listeners unfamiliar with the blind test situtation fail to detect even quite big differences.

It works the other way round too; trained listeners (that means trained under the specific conditions of the blind test choosen) can reach astonishing levels of sensitivity.

Normal statistical analysis is conservativ by definition, so is first of all interested in avoiding a false positive, but as we are testing to get better confidence in our knowledge about what can be heard, we should be interested in avoiding a false negative too.

In this respect Les Leventhals articles were quite informative back in the 80ies, but it was just basic stuff that audio people were unfortunately not aware of in those days.

And because a audio blind test is as much a test for the listener as it is a test for the maybe existing difference, it is so important to include controls to get more information about the usefulness of the test results.

Positive controls are differences that should be detected in the blind test; if incorporated on different sensitivity levels that gives you information about the sensitivity a specific listener reaches.

Negative controls are non existing differences that should be reported as non existent by the test results.

Jakob
 
#828
SleeperSupra said:
I was on another audio forum and was told that:

"The science behind hearing and amplifiers tells us that the person making the claim that he hears a difference between two (properly functioning) amplifiers does in fact believe this claim, but that there is in fact no audible difference. That's not a bad thing, the reasons are pretty well understood (at least among people who study psychoacoustics), they just have nothing to do with the amplifiers. IOW, if you think you can hear the difference between two normally functioning amplifiers then you deluding yourself, it's a common delusion, but a delusion none-the-less..."

"The basic theory is that two amps will sound alike unless:

1) One is playing louder than the other.

2) There are audible differences in frequency response (which can often be caused by impedance mismatches, particularly in the case of tube amps).

3) One is clipping.

So if you hear an amp as more forward or warmer (whatever that means to you), it has to be caused by one of the conditions above.

Now, here's the thing: Among modern solid-state amps driving typical home speakers, it is very rare for conditions 2 and 3 to hold. As for #1, that has nothing to do with the character of the amps, and everything to do with where you set the volume control. Tweak it a little, and both amps will sound the same.

So your next question is, OK, if that's all true, why do I and other audiophiles hear differences between amps? Three reasons:

1) You're comparing without matching output levels, which has to be done very precisely (i.e., your ears and/or a Radio Shack SPL meter aren't enough; you need a voltmeter measuring the signal at the speaker terminals).

2) You aren't comparing the two side by side, but are relying on your long-term memory of the sound of one of them. Our long-term memory of subtle sonic differences is really poor.

3) You are subject to what the psychologists call "bias," which simply means that your hearing perception is influenced by other factors--specifically, other things you think or know about the amps. If some salesman once told you that Brand X tends to be warm, that's liable to affect how you hear Brand X. Can't be helped, as you're only human. That's why scientific listening comparisons are always done blind."


I do not believe this to be true as I have owned over a dozen amplifiers and each one had a unique character to its sound.

What is your take on this subject?
Click to expand...

My opinion is that there is no real argument above. A real argument would be that there is a 10% of occasions in each of our individual life long audio experience that it happened for conditions to match and we got surprised by all amps sounding the same. If it does not happen in life, it is not happening. Period. Mathematical simplicity is wonderful only if it is not simpler than the problem. Then its simplification and not simplicity, and serves only to derail the train to a solution.
 
#829
There all all sorts of confounding issues when doing blind testing that involves people and auditory perception.

The simplistic run down quoted in the previous post(s) is to this sort of test what a kitchen table radio (do people still have these?) is to a high quality, high resolution hi-fi system.

It is true what is being said, in the main.
But the conclusion drawn and inserted into the text that there are not (audible) differences is an unwarranted proposition.

Unless and until someone does some ABX or other blind tests that covers not ONLY the points mentioned above, but a wide range of points beyond those (that I've mentioned earlier... read back), there is no way to GENERALIZE beyond the results of a specific test condition to all situations or all amplifiers.

The biggest problem with any test, no matter who sets it up or how careful they are is the potential for one or more sources of masking. Masking is sufficient to render all results invalid.

Without the full and proper documentation of the test conditions and controls of the test conditions, it is A) imposssible to recreate the tests, B) impossible to know if the results have any general applicability.

A very simple example of a basic control is to demonstrate in the test that something sounds different and what and why it does... that's just a very simple basic necessity, eh?? (that's at least a test of "degree of sensitivity" in some regard... albiet a rather nonspecific one) Hasn't happened yet, don't hold your breath waiting?

The responsibility for "proof" lies solely with those who carry out "tests" that they publish in peer reviewed journals, not with those who make this or that claim, one way or the other. Be that here or in other forums, on websites, or in magazines, the only responsibility for being thorough and truly scientific lies with those who have decided to do tests and publish the results in Journals. Everything else is posturing and propaganda.

Enough already.

_-_-bear
 
#831
Re: Re: Do all audio amplifiers really sound the same???

salas said:


My opinion is that there is no real argument above. A real argument would be that there is a 10% of occasions in each of our individual life long audio experience that it happened for conditions to match and we got surprised by all amps sounding the same. If it does not happen in life, it is not happening. Period. Mathematical simplicity is wonderful only if it is not simpler than the problem. Then its simplification and not simplicity, and serves only to derail the train to a solution.
Click to expand...


Well, I'm not saying all amps sound the same. I have no proof they would. It seems reasonable to expect cases where the differences between amps would be so large as to be audible. The issue how do we find out whether two amps really do (or do not) sound the same?

The post you quoted above again shows the inconsistency of thought of many people. First there is a whole dissertation of possible causes of undetected differences in dbt etc, and then he says:

"I do not believe this to be true as I have owned over a dozen amplifiers and each one had a unique character to its sound."

Sorry, but that was just the thing we were trying to (dis)prove! This is circular reasoning at best.

Then in the first part of the quote it is stated:

"The science behind hearing and amplifiers tells us that the person making the claim that he hears a difference between two (properly functioning) amplifiers does in fact believe this claim, but that there is in fact no audible difference. That's not a bad thing, the reasons are pretty well understood (at least among people who study psychoacoustics), they just have nothing to do with the amplifiers. IOW, if you think you can hear the difference between two normally functioning amplifiers then you deluding yourself, it's a common delusion, but a delusion none-the-less..."


It is correct that if a person believes to hear any difference (whether it actually is there or not) does so entirely honestly and genuine.
If you remember the example of the Bosnia crisis, where neighbours of long standing friendship turned into deadly enemies, they honestly and with conviction believed that those earstwhile friends were real enemies that would kill them if they wouldn't do that first. In the theory of Cognitive Dissonants, peoples (changed) convictions are real and genuine.


It is also fully correct that these phenomena have indeed nothing to do with amplifiers and audio per se, but with almost all cases in life where people have to chose between courses to take, things to believe, advice to follow. The saving grace of audio is that nobody dies if they erroneously believe that there is an audible difference and there isn't any. But if you take drugs, people *could* die if they erroneously believed that a certain drug would cure them instead of making them more ill.
That's the reason for the extensive db tests done on medicine before they are released to the market.


I also feel that the term 'deluded' is unneccesarily negative. If a person who receives a placebo instead of a medical substance reacts as if he/she actually received the medical substance, deluded? Or is it just one of those typical human things in our wiring we have to live with?


Jan Didden
 
#832
Pan said:
scott,



Well theres hardly a need for such an expert to point out something as obvious. However the illusion of a 3D space between and behind the speaker set up can be frightingly good. For a full soundfield to be reproduced you need many channels. I'm about to start out with eight channel recordings to see what it gives.



Yup, low noise is important it seems and I assume you are thinking on dynamic noise primarly and not the static background noise.


/Peter
Click to expand...

Yeah basicly, static noise goes without saying but the work people have done on switching and modulation effects in power supply rectifiers, and RFI elimination in general has been very important. Did you ever notice how many hi-res THD spectra show 50/60Hz (and harmonics) sidebands? Just wait until Homeplug power line networking takes off.
 
#833
Its just that solid and expensive science has other priorities than audio. The phenomena can be explained. See Earl Geddes solving horn issues. Former math proved inadequate. Almost everybody was hearing horn sound coloration all the time though. If you asked the classic school (all the horns and drivers at work today), they could argue that it was nothing wrong with the established horn math and practices. Before Earl. And if there is vested interest...he will have to prove that he is not an elephant.
 
#834
If there is a placebo effect so strong that it can fool even the most experienced, which I doubt very much, then it will actually work in favour of "the same" people
You see, they will always have a strong belief that they have an amp as good as any, no matter whether it really is good or not, where as "the different" people always will be more critical and expect that there is an even better amp out there
"The same" people will be content and satisfied, or at least accept things as they are... "the different" people will be in constant search fore better
 
#836
The problem is that there are lots of listeners and manufacturers that actually "suffers" from strong placebo like effects. No point in denying the voodo and mumbo jumbo in the biz.

So in a way I can understand the scepticism that I see from some of my fellow forum members all over the world. 🙂

However I can't accept the attitudes that some are showing.. a little humbleness seldom does any harm.

And remember.. there was a time when the earth was flat and bacteria did not exist. 😉

BTW good posts from Jacob and others explaining the pitfalls and possible problems with testing.


/Peter
 
#838
Sure, there is a placebo effect, i have never denied that...but we are some that arent impressed by size or fancy design😉 but I do know intelligent, skilled and experienced people who believe very strongly that diy cannot ever be any good, no matter what...so yes, I do know about placebo effects :clown:

btw, I believe what I personally hear...not what other people hear
 
#839
Bear, there was a bibliography several pages back. I'd particularly point you toward Lipshitz, Olive, and Toole's papers. They unfortunately are not free, but if you're an AES member, they're not too expensive, and of course, there are engineering libraries.

Don't believe the cartoon versions discussed on the Internet- it's worth reading the papers in their entirety. You'll be surprised at what they conclude and don't conclude. And you'll be impressed with the care of their methodology and the level of attention to detail.
 
Status
Not open for further replies.
Top Bottom