Double Blind tests *did* show amplifiers to sound different

[amirm]

In many online debates, position taken by some that when amplifiers are used that have flat frequency response and low distortion that no double blind tests have shown them to sound different.

Well, I managed to dig up a 32 year old test that says otherwise. What is fascinating is that one of the authors/testers was no other than [the late] Arny Krueger, who was one of the earliest advocates of ABX type testing.

The article is actually quite short. What I am about to show you is about half of it and contains all pertinent portions. I am happy to field and answer questions that may be in the rest of the article.

So let's start at the top. Here is the title of the 2-page article:

What a headline, no? It says they have "proven" that amplifiers sound different! You almost don't need to read any more but for completeness let's go into it.

Here are the amplifiers in question:

So these are very powerful amplifiers. We are not talking a tube SET amplifier or some other amp that runs out of juice at 20 watts. We have very powerful amplifiers clocking at over 200 watts at 4 ohms.

This being the pre-CD era, a turntable was used. Here are the details of that and the speakers used:

So how did they do? Here are the votes for the three tracks tested:

We see why they say "prove" as the two tracks show statistically that the results are not due to chance.

So the conclusion is as expected:

So what caused the difference? It might be this:

No data is presented on what would happen if the volume was turned down. Or how the above was determined.

So there you have it. "Proof" that amplifiers do sound different in double blind tests .

EDIT: updated the article with local version of the text as the shared links had disappeared. Also cleaned up the language a bit.

 

[tomelex]

Well we all know by now that simple THD and FR are not enough to characterize amplifiers. The harmonic spray and output impedance are two things that easily influence differences in sound between amps of the same type of output topography. Amps act different driving real loads vs an 8 ohm resistor load bank. There are a lot of technical reasons but those two are the biggies

 

[amirm]

Indeed. I think one key area not tested is the effect of the protection circuit and power supply lag during peaks. Combine this with feedback and it gets ugly really fast.

Here is my Pioneer Elite AVR (class D)

This one I think is Onkyo AVR (Class AB)

These are clippings into resistive load. As you mention Tom with a real speaker, the phase and low impedance can work together to make the load look far more difficult for bass frequencies causing excursions into and out of the above waveforms.

Arny in a conversation in HA forum about this test said that they were comparing amplifiers when they heard a buzzing sound on one of the amps and tracks and decided then to do this blind tests. And of course they heard it. It could have been oscillation or one of the above factors.

 

[Phelonious Ponk]

Can you explain what those two charts show, to the ignorant and unwashed?

Tim

 

[amirm]

Sure . The amp was driven with a pure sine wave into a resistive "dummy" load. Then the volume was turned up until a lot of distortion occurred. What the graphs show is that the typical picture we have in mind of a pretty clean, clipped flat top is not what happens. Instead we get very odd shapes.

The odd shape occurs because when you reach clipping, the protection circuit kicks in to limit current and save the output transistors. This makes the feedback circuit upset whose job is to make the input the same as output. So in turn it turns up the gain to counteract what the protection circuit is doing!

As the two fight another implicit feed look which is the power supply. In almost all amplifiers, the output stage runs unregulated meaning it starts with a high voltage but then sags as the load increases. Feedback in the amplifier makes it pretty immune to causing distortions. At clipping we have a new mess on our hand because now the power supply voltage goes up and down depending on what fighting is going on between the feedback loop and protection circuit.

In lay terms, there are three subsystem all working against each other with no coordination. The result is the twisted shape of the clipping and also oscillation which is shown in the first AVR.

The above is one of the reasons some folks don't like feedback. Feedback works great until you arrive at the above situation. It then works against you and makes the distortion worse, not better. So clipping that perhaps was benign is no longer so.

All of this can be avoided if you have ample amount of power available to you. Then you don't worry about the peaks.

Hope this makes sense. Please ask if it does not.

 

[Liberal Democracy]

yea that makes sense, thanks tim for asking what i might of but for fear of looking turbotarded (well more turbotarded)

sounds similar to banking/trading regulation

 

[tomelex]

This is why clipping indicators are so valuable, the so called hi end abandoned them because their (at the time) lower powered gear spent a lot of time in clipping. Tube amplifiers, for example, SET amps without feedback, clip less harshly, but still clip, however the wave is more rounded and thus in simple terms less harmonics created and less odd order harmonics created. Its a generalization but still factual

 

[amirm]

I agree. Love to see the lights if they are well implemented (not all are). I have them on my tape deck and it is nice to see it go nuts and listen to distortions caused by peaking that way.

 

[Blumlein 88]

I know this is a year plus later, but just noticed this thread. I feel it worth adding the following information.

I somehow missed this when posted. I am shocked that the speakers were characterized as about 5 ohms of flat impedance. Somebody didn't do due diligence.

I owned Acoustats with that interface. The Acoustat Twos which is what they used if I understood them rightly. I setup and measured the impedance of them with a sine generator, inline resistor and meters to know phase and impedance from 10hz-25 khz. My curve looked essentially the same as what I obtained doing this with Soundlabs. Also essentially the same as the Stereophile measures for Soundlabs shown below.

Look at the impedance curve of Soundlab ESL's, the same basic shape as Acoustats. The peaks are in slightly different places, but the same general trend. Both brands used the same xfmr arrangement (going by memory) a 200:1 transformer for the low end, and 50:1 for higher frequencies. These have high level crossovers blending the output onto the panels. The 7-800hz peak is where both halves are blended together. Both Soundlabs and Acoustats will drop to or depending upon the brilliance setting (essentially an inline rheostat) below 1 ohm at high frequencies. Output impedance will alter amplifier balance perhaps more than 99% of all speakers.

Quoting from JA's measurements of the Soundlab A1:

This measurement necessarily includes the resistance of the speaker cables DO was using (TARA Labs Rectangular Solid Core); as the margin of error in my measurement is probably around a fraction of ohm, it's quite possible that the A-1 presents amplifiers with a complete short circuit above the audio range. Fundamentally, therefore, in electrical terms, the A-1 with its BC set to maximum is nothing more than a large capacitor. If this isn't hell for a power amplifier, I don't know what is (footnote 1). No wonder DO's amps were choking.

The footnote tells how they tried using Fourier OTL's, but the output tubes immediately turned cherry red with a very low signal level. These are good speakers to differentiate between amps. Many would sound very different with these which would sound identical with most loudspeaker loads. At 20 khz you are seeing about an ohm and a 75 degree phase angle.

My Wyred4Sound ST500 plays my Soundlabs with effortless aplomb in case anyone can use that information.

 

[amirm]

Indeed my belief is that the difference in amplifier sounds occurs at the limits of their output. It is then that each amp acts very differently than others.

 

[Cosmik]

This experiment was a perfect demonstration of how 'science' is carried out using a rigorous, neutral methodology and so on, but the only bit that anyone looks at is the headline and conclusions, and these are spun out of thin air to reflect whatever agenda the writer wants to push. In this case, the headline could have said "Amplifier sound differences may be due to clipping" but this wouldn't have attracted as much attention.

 

[Jinjuku]

The paper lays out the conditions that Amps can start sounding different. The title is a bit 'Read Bait' but they are clear in their conclusions.

Nothing to see here folks, move along.

 

[Cosmik]

The title is a bit 'Read Bait'

... repeated (almost) in the conclusions

but they are clear in their conclusions.

If a conclusion is clear does that automatically make it correct, relevant and justifiable?

 

[amirm]

So a bit of history on this. For years I was engaged with arguments with one of the authors of that paper. By chance one time he mentioned the title of this article. But it had come out years and years ago in a magazine and there was no record of it anywhere online. In yet another argument, he pointed to an auction that someone was selling this specific issue. I paid my $20 and was surprised to see so much detail in there that the person I was arguing with had not disclosed.

 

[Jinjuku]

... repeated (almost) in the conclusions

If a conclusion is clear does that automatically make it correct, relevant and justifiable?

Usually it makes it open to independent testing and either verified or debunked.

 

[Cosmik]

Usually it makes it open to independent testing and either verified or debunked.

The experimental results are open to independent testing and verifiability, but the conclusion is not in a suitable 'scientific format'.

For example, in an experiment based on listening to two amplifiers playing one second of noise, where one listener cannot distinguish between them, and the conclusion contains the words "All high quality amps sound the same", then this conclusion is neither verified nor debunked by independently repeating the experiment. Only the results are.

 

[Blumlein 88]

This is why clipping indicators are so valuable, the so called hi end abandoned them because their (at the time) lower powered gear spent a lot of time in clipping. Tube amplifiers, for example, SET amps without feedback, clip less harshly, but still clip, however the wave is more rounded and thus in simple terms less harmonics created and less odd order harmonics created. Its a generalization but still factual

When I first acquired Acoustat Twos I owned a Carver receiver. The seller said it would never do. I asked if I could try it. It played those nasty Acoustats very well. The seller was having trouble believing it. I did get better amps later, but it worked fine for 6 months.

One of my pet peeves has been how many people I knew that regularly clipped their amps. And how this is why some of the beliefs about what amps were good or bad or sounded different came from. Often all that was needed was turning down the volume just a bit or buying a proper sized amp to start with. Exotica wasn't required.

At one point between amps I had a McIntosh 752 amp. 75 wpc which wasn't really enough for the Acoustats. It however had Power Guard. It was an opto-coupled anti-clipping circuit. Lights would flash and dynamically reduce signal level to prevent clipping and harsh distortion in 1 millisecond. I think they claimed it prevented distortion from ever reaching .5% with that. Though not my normal listening I did at times run the level way up just to see. It really worked. You could run level up, the lights flashing on and off furiously, the sound would be compressed just like a compressor effect in studio processing of recordings. It never got nasty at all. At an extreme it was very loud, very compressed. So much gear would do well to have such a circuit for clipping conditions. This circuit was not in the signal path either. Those lights are what engaged a reduction of voltage across the transistors. Protected the output transistors and listener's ears.

http://mcc.berners.ch/power-amplifiers/MC752.pdf

 

[RayDunzl]

Amplifiers sounding different?

I seem to remember having that sensation. Maybe more than once. Haven't you?

They certainly can be constructed differently.

Blind tests conducted under strict scientific supervision?

No, I haven't been there. My ears aren't certified.

---

But if you're a hobbyist, and what you hear floats your boat, and won't break the bank, and makes you feel good, go for it.

I am, it did, it didn't, and it still does, and I did. Actually, twice in a row ("Why wait another 15 years for the next step?", I said to myself, a week into the first upgrade).

Or use it as a reference (in your possibly fallible long term acoustic memory).

---

Though I might be a little more suspicious of myself now. Surely that will affect my next major purchase, if there is one.

---

Just don't tell anyone.

Certainly not here. Unless, of course, you have the right sets of measurements.

Or at least be circumspect about it. Smile and nod approvingly at the dogma. Practice that. Good. Again. Good.

Maybe pick an avatar with a slight sneer to give hidden hints about all your posts. Maybe someone with some quotable quotes:

• "All my life I wanted to look like Elizabeth Taylor. Now Elizabeth Taylor looks like me". - Glenn Milstead

---

Elsewhere, knock yourself out.

Pick another name and avatar, just to confuse things.

 

[fas42]

...

Maybe pick an avatar with a slight sneer to give hidden hints about all your posts. Maybe someone with some quotable quotes:

• "All my life I wanted to look like Elizabeth Taylor. Now Elizabeth Taylor looks like me". - Glenn Milstead

---

Elsewhere, knock yourself out.

Pick another name and avatar, just to confuse things.

... "Come here to have fun, be ready to be teased and not take online life too seriously."

 

[Sal1950]

Knowing what we do today about "proper" amp, cable, speaker designs I was and mostly still am of the opinion that all modern, properly designed amps should sound alike (or similar in the extreme}. But the one thread here that most clearly opened my eyes as to how far some modern gear can stray from what most would consider "properly designed" were the Measurements on amplifiers with speakers and cables connected posts.
Highlighting to me how strange things can get was the situation with member cjf's Magico S3 speakers and his cable/amp interfaces. Starting at post #49 I followed the tech talk as best I could, it became clear to me that even knowing everything we do today, there are a number of manufacturers not following what I would see as a modern enlightened convention? Maybe this also highlights how the split between hard scientific and boutique designs has landed us?
YMMV or I'm showing my ignorance?