All power amps “good ones” sound the same.

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
JohnRoberts said:
I am not sure of your point(s), and what "difference" your tests would reveal.

My point is that a test can be designed to confirm bias. Whatever that bias is.


There are comparison strategies that are more revealing of small differences, like null tests.

Like Abbey said if you use a resistive load everything will measure great.

Try driving a difficult load in the real world. Not so much.

Only listening tests will tell you that.
 
Gold said:
My point is that a test can be designed to confirm bias. Whatever that bias is.
well controlled double blind listening tests with enough trials to realize statistical significance can determine if there are actual audible differences or not. The double-blinding thwarts expectation bias.
Like Abbey said if you use a resistive load everything will measure great.
not always... but most modern amplifiers are very good.
Try driving a difficult load in the real world. Not so much.
I spent decades doing exactly that... You should have seen the "difficult" loads that Jack Sondermeyer (RIP) used on his test bench when checking out amplifiers. He released the magic smoke from inside more than one competitor's amplifier doing stress testing.

Only listening tests will tell you that.
I am far less optimistic about listening tests. I only know with any certainty what I can't hear. I can measure differences that I can not hear, I never heard a difference I couldn't measure, while some transient events are more difficult.  (I rolled a custom variable tone burst generator to help parse out transient artifacts).

As I already shared I went to lengths in that amplifier test last century to load the amplifiers with actual loudspeakers (IIRC I may have loaded them in parallel down to 2 ohms nominal for some of the trials), and then cranked them up loud. If you think you can hear subtle differences between big amps driving sound reinforcement speakers at high volume, you are a better man than I...

I performed my critical listening to the null signal in a relatively quiet room nearby... The sound proof room wasn't completely soundproof but adequate. 

In my experience if we determine that there is a difference, we can then dig deeper on the test bench to quantify what that difference is. If you can measure it you can manage it and design out any such flaws. Indeed inappropriate current limiting can be audible and I suspect one of the accidental benefits of the popular early (Hitachi) lateral mosfet amplifiers, is that some of the amp designers punted on using any current limiting at all because lateral mosfets enjoyed a positive tempco preventing thermal runaway. Further the lateral mosfets had high gate capacitance that could compromise marginal drive current in output stage drivers. This softer high end, and lack of current limiting may have sounded better than mediocre bipolar amp designs of the day.

Likewise some inexperienced exotic loudspeaker designers could be optimistic about amplifier drive capability. At Peavey since we designed both loudspeakers and amplifiers it was hard to ignore that interaction. We would hear quickly about any mistakes we made from the dealers/customers. 

JR
 
JohnRoberts said:
well controlled double blind listening tests with enough trials to realize statistical significance can determine if there are actual audible differences or not. The double-blinding thwarts expectation bias.

I meant the bias of the test designer, not the subjects. You want a test that shows no difference? I can make that test. You want a test that shows a difference? I can make that test too.

As I already shared I went to lengths in that amplifier test last century to load the amplifiers with actual loudspeakers (IIRC I may have loaded them in parallel down to 2 ohms nominal for some of the trials), and then cranked them up loud. If you think you can hear subtle differences between big amps driving sound reinforcement speakers at high volume, you are a better man than I...

Who said PA speakers at high volume? The example I gave was of speakers that straddle the audiophile and pro audio world. B&W 801 and 802. The reason I used those as an example is that I have heard those speakers in multiple rooms with multiple amps. I've even heard a good portion of them in various Fran Manzella Designed rooms. I've made better/worse, same/different listening decisions every day for over 30 years. People still pay me to do that.

In my experience if we determine that there is a difference, we can then dig deeper on the test bench to quantify what that difference is. If you can measure it you can manage it and design out any such flaws.

It is difficult to measure dynamic behavior and that is where the differences lay.
 
Gold said:
I could easily design a test where you couldn’t tell the difference between two amplifiers. I could also design a test where you easily could.

Really? perhaps you should share it, notice that I said that the amp should not be clipping, nor that anything inherently wrong be with them, like a bad design, faulty part,  limited frequency response, etc... Also, no difference in level should be present. If you could, easily, design a double blind test like that, you should do it, you would do a service to the audio industry. Now, if you are going to just make an A minus B test and claim there is a difference, well, I don't think that is very useful or the point of this test, the point is if the user can hear a difference under normal listening conditions.
 
user 37518 said:
Really? perhaps you should share it,

Sure, pick two decent amps and drive an Auratone. You won't hear much if any difference. Do the same with a B&W 801. You will hear a difference.
 
A lot of Peavey PA gear did tollerate 2ohm loads , you'd normally only opperate at 4 ,but in  pinch you could patch a set of 8ohm side fills off the main amp no worries, 

Three Bose 802's (with bins and system EQ)off each side of a Peavey CS-800 , still stands up against many modern  Array systems for vocal carry.
 
Gold said:
Sure, pick two decent amps and drive an Auratone. You won't hear much if any difference. Do the same with a B&W 801. You will hear a difference.

I'm not sure you will hear a difference with a B&W, in the blind test link I posted, they used ATC speakers, people did not notice a difference. Your test is based on supposition, not on real fact, you are assuming people will notice a difference, I encourage you to do an ABX double blind test with a pair of B&Ws and see if in fact people notice a difference. P.S. it has been done before... Same thing with converters, I can post a reference published in the AES, people couldn't hear the difference between a resampled audio and the original high sampling rate audio, even after multiple generation convertions, again, this is based on the scientific method not on audiophoolery.
 
user 37518 said:
I'm not sure you will hear a difference with a B&W, in the blind test link I posted, they used ATC speakers, people did not notice a difference.


Just in case you missed it from the above post. And I've done the Auratone test too.


"The example I gave was of speakers that straddle the audiophile and pro audio world. B&W 801 and 802. The reason I used those as an example is that I have heard those speakers in multiple rooms with multiple amps. I've even heard a good portion of them in various Fran Manzella Designed rooms. I've made better/worse, same/different listening decisions every day for over 30 years. People still pay me to do that.


I absolutely hear differences between amps. I don't want to hear the 'it's not a real A/B/X test" excuse either. I know how to listen. Let me know who has the funds to set up a valid ABX test and I'll be there. If I can hear a difference I don't need people to tell me.

I use ATC SCM100's in my mastering room but they are powered.

I have no idea how resampled audio relates to power amps.
 
Has anyone here participated in a scientifically valid double blind A/B/X test for anything?

It’s very difficult and expensive to set up properly.

I know I haven’t. That’s why I think most of the time people who say that have no idea what they are talking about.
 
Tubetec said:
A lot of Peavey PA gear did tollerate 2ohm loads , you'd normally only opperate at 4 ,but in  pinch you could patch a set of 8ohm side fills off the main amp no worries, 

Three Bose 802's (with bins and system EQ)off each side of a Peavey CS-800 , still stands up against many modern  Array systems for vocal carry.
Peavey amps would not drive 2 ohms all day long, but I heard more than a few anecdotal reports about customers in a bind driving them down to one ohm or less to make it through a gig. The show must go on. 

JR
 
Gold said:
Has anyone here participated in a scientifically valid double blind A/B/X test for anything?

It’s very difficult and expensive to set up properly.

I know I haven’t. That’s why I think most of the time people who say that have no idea what they are talking about.
who me?

:)

That's why people don't do them... they are hard.

They especially hate being told that they didn't consistently hear a difference.

I recall a guy in Michigan making and selling commercial ABX test boxes for blind testing (David Clark?) back in the 80s but it fell out of fashion, because the big business marketing programs didn't find it profitable to test scientifically.
===
I've shared this story before, but back in the late 80's/early90's A recording magazine performed a double blind listening test comparing multiple popular studio monitors. At the time Peavey's AMR division had a series of studio monitors targeting the voicing of several popular monitors (NS10, etc).

Long story short, the double blind listening test was performed in the bay area, and more than one studio puke withdrew and declined to allow his name being associated with the test after he was informed that he had personally picked the Peavey (AMR) monitor as best.

JR
 
Gold said:
Just in case you missed it from the above post. And I've done the Auratone test too.


"The example I gave was of speakers that straddle the audiophile and pro audio world. B&W 801 and 802. The reason I used those as an example is that I have heard those speakers in multiple rooms with multiple amps. I've even heard a good portion of them in various Fran Manzella Designed rooms. I've made better/worse, same/different listening decisions every day for over 30 years. People still pay me to do that.


I absolutely hear differences between amps. I don't want to hear the 'it's not a real A/B/X test" excuse either. I know how to listen. Let me know who has the funds to set up a valid ABX test and I'll be there. If I can hear a difference I don't need people to tell me.

I use ATC SCM100's in my mastering room but they are powered.

I have no idea how resampled audio relates to power amps.
"I know how to listen" that argument just reinforces what I was saying, you cannot provide a valid scientific test that clearly show the differences. You said you could do that and your entire argument is based on "I know how to listen"...

And you didn't see the link I posted in which they CONDUCTED an ABX blind test and statistically no one could tell the difference between a Behringer amp and an expensive one through ATC speakers.

http://matrixhifi.com/ENG_contenedor_ppec.htm

Ricardus said:
Yeah no. Our ears are the weakest link in the test chain, and our lizard brains lie to us.

Double blind or nothing.

I agree, it must be double blind ABX. Otherwise its just what audiophools do, they listen to expensive amps and confirmation bias makes up for the rest.
 
Amplifiers which don't handle "difficult" or complex loads at high power, how well do they fare at levels you would expect at home listening or in a studio? I rarely see any home stereo amps that claim to be less than 100W. I could be wrong, but I would expect almost any 100W amp to drive any "hi-fi" speakers no matter how exotic crossover it has without any trouble at home listening levels.
 
Heikki said:
Amplifiers which don't handle "difficult" or complex loads at high power, how well do they fare at levels you would expect at home listening or in a studio? I rarely see any home stereo amps that claim to be less than 100W. I could be wrong, but I would expect almost any 100W amp to drive any "hi-fi" speakers no matter how exotic crossover it has without any trouble at home listening levels.

Class AB amps usually get hotter at around 1/3 their rated power and Class A amps dissipate the least power at full power.
 
user 37518 said:
"I know how to listen" that argument just reinforces what I was saying, you cannot provide a valid scientific test that clearly show the differences.

Of course I can’t. I’m not set up to do it.

I saw the link. I’ll read the AES paper. I’m sure the setup wasn’t trivial. Like I said. I’m confident that if the test I suggested was set up double blah blah it would prove my point.

You said you could do that and your entire argument is based on "I know how to listen"...

You don’t know how to listen? You’ve never made an audio decision by listening? No one could tell the difference between a victrola and a live musician in a double blind blah blah at the turn of the 20th century. I don’t think the result would be the same today. Are our ears better today? No, we are familiar with the artifacts of a victrola which was new at that point. You need to be a trained listener. I am not making the golden ears argument.

Now granted I thought “different amplifiers sound different” was uncontroversial. Apparently I was wrong.
 
Gold said:
I could easily design a test where you couldn’t tell the difference between two amplifiers. I could also design a test where you easily could.

Gold said:
Of course I can’t. I’m not set up to do it.

So I guess its not that easy to do a scientific test that easily proves that two amps sound different...
 
I'm with Paul on this one and never understood the rigid everything must be double blind ABX or else there is no difference stance. What often separates professionals is the ability to make quick "instinctive" decisions that are correct a high percentage of the time. Their judgment shouldn't be so easily discounted.

I also don't get why people think confirmation bias exists only on one side.  There are plenty of people who don't want to hear a difference.
 
user 37518 said:
So I guess its not that easy to do a scientific test that easily proves that two amps sound different...

No. I never said it was. The AES paper proved that the equipment in that situation produced that result. I also said I could design a test to show no difference or a difference between amplifiers depending on the equipment I picked. Whoopee.
 
Back
Top