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 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,

He passed a couple of years ago...😢

I actually have a limiter he made back in his earlier college days and he was always helpful the few times I had questions. Very passionate about sound...
RIP
 

Attachments

  • JAES_V66_10_PG863.pdf
    34.4 KB · Views: 4
john12ax7 said:
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.

There is a difference between art or the creative process, for example when doing a mix or producing a song, and another is the difference between what can be proven or not. In the medical industry they realized about this very early on, thats why they don't tell patients they are taking the placebo, that influences not only their actions but some of them actually feel better thinking they are taking the real thing, also, the research doctors don't know which patients are taking the placebo or the real drug because it also influences their findings or conclusions.

Thats the entire logic behind double blind testing, if professionals are so confident on what they can hear, then there shouldn't be a problem with double blind ABX test, the problem is that in the conditions I've been explaining on all my previous posts, people can't hear it, and thats why they try to disprove it with many (pointless) arguments.
 
I have to attest to the fact that B&W 801's are hard to drive for some reason. I have a pair and found that out.

I'm a big ABX test guy and have done many with converters and sample rates. In which I came to a conclusion that I, nor anyone else at the studio I work at, could hear a difference between pretty much anything when the test was set up to a high degree of level accuracy, and truly blind. I've gone on and on about this before here so I wont go into detail.
I say that just to give a reference point for what follows.

My direct experience with the B&W's and two different amps is that I heard a BIG difference between a Hafler p7000 and an Ashly FTX1501 (175W/side). The Halfler is a monster linear amp that can do 350w into 8 ohms a side. The Ashly was a professional sound reinforcement amp so not some weak piece of junk. But definitely not as powerful.

I switched back and forth quite a few times and the difference was obvious. No it wasn't blind and the switching method took time to plug and unplug amps. Not scientific at all BUT... (I won't go into all the audiophool words here) it was "better" (The Hafler) by a long shot. I know the Ashly was only 175w a side but I was listening at volumes that shouldn't have been stressing that power at all. (I want to add, I had recapped the Ashly a month or two before this happened)

Now as far as bias (what I wanted to hear) is concerned, The Hafler turned out to be nosier mechanically because of the cooling fan which revved up really high as it started to heat up. It was annoying and eventually I went back to the Ashly amp in hopes the difference wasn't as noticeable as I thought it was. Nope, it was worse. I went back to the Hafler.

I'm not trying to prove anyone right or wrong but just sharing an experience. Perhaps an Ashly 350W/side amp would have stacked up better.  I too am a mastering engineer and "listen" in an acoustically treated room everyday for a living. I don't consider myself to have any type of "golden ears" at all. Just pointing out that I have a little bit more cred than most casual listeners.

I kept the Hafler until I got a couple Hypex 400W class D modules and built myself a nice quiet amp. the difference between the Hypex and Hafler was small enough that I can't be sure there was any. I just know I like the class D amp because it runs cool and doesn't make any noise.

Maybe it was purely a wattage thing and maybe it was circuit design. I would say its the latter.

But I think there should be room for differences in power amps more so than converters or sample rates. There's more physicality to it. It seems there are too many variables to make any statements about it in either direction.

We did however do a "as close to you can get to a scientifically valid A/B test" in a studio tracking room between Bryston linear amps and Hypex class D amps. It was definitely interesting. Too much typing to explain that whole thing...



 
bluebird said:
I have to attest to the fact that B&W 801's are hard to drive for some reason. I have a pair and found that out.

I'm a big ABX test guy and have done many with converters and sample rates. In which I came to a conclusion that I, nor anyone else at the studio I work at, could hear a difference between pretty much anything when the test was set up to a high degree of level accuracy, and truly blind. I've gone on and on about this before here so I wont go into detail.
I say that just to give a reference point for what follows.

My direct experience with the B&W's and two different amps is that I heard a BIG difference between a Hafler p7000 and an Ashly FTX1501 (175W/side). The Halfler is a monster linear amp that can do 350w into 8 ohms a side. The Ashly was a professional sound reinforcement amp so not some weak piece of junk. But definitely not as powerful.

I switched back and forth quite a few times and the difference was obvious. No it wasn't blind and the switching method took time to plug and unplug speakers. But I did move the speakers to the same positions. Not scientific at all BUT... (I won't go into all the audiophool words here) it was "better" (The Hafler) by a long shot. I know the Ashly was only 175w a side but I was listening at volumes that shouldn't have been stressing that power at all. (I want to add, I had recapped the Ashly a month or two before this happened)

Now as far as bias (what I wanted to hear) is concerned, The Hafler turned out to be nosier mechanically because of the cooling fan which revved up really high as it started to heat up. It was annoying and eventually I went back to the Ashly amp in hopes the difference wasn't as noticeable as I thought it was. Nope, it was worse. I went back to the Hafler.

I'm not trying to prove anyone right or wrong but just sharing an experience. Perhaps an Ashly 350W/side amp would have stacked up better.  I too am a mastering engineer and "listen" in an acoustically treated room everyday for a living. I don't consider myself to have any type of "golden ears" at all. Just pointing out that I have a little bit more cred than most casual listeners.

I kept the Hafler until I got a couple Hypex 400W class D modules and built myself a nice quiet amp. the difference between the Hypex and Hafler was small enough that I can't be sure there was any. I just know I like the class D amp because it runs cool and doesn't make any noise.

Maybe it was purely a wattage thing and maybe it was circuit design. I would say its the latter.

But I think there should be room for differences in power amps more so than converters or sample rates. There's more physicality to it. It seems there are too many variables to make any statements about it in either direction.

We did however do a "as close to you can get to a scientifically valid A/B test" in a studio tracking room between Bryston linear amps and Hypex class D amps. It was definitely interesting. Too much typing to explain that whole thing...

You just said it, with an ABX double blind no one could hear a difference, with a sighted test you could hear a difference. Point proven.
 
bluebird said:
Yup, that's why I typed that whole post, just to prove your point. :p

What I get from your post is that you made an ABX test and no one in the studio could hear a difference, but with an unscientific sighted test, you prefered the Hafler amp to the Ashly amp, which you later replaced with a class D because it runs cooler and makes less fan noise. An opinion that should count more because you are a mastering engineer.
 
The ABX test was with AD/DA converters. I was just explaining that to show I'm a fan of the scientific method. It had nothing to do with the amp test.

Its all good, I was just telling a story, not trying to prove a point or anything.
 
P.S. Being an audio engineer doesn't mean that much, in fact, can't remember exactly which study (double blind ABX) the audio engineers performed worse than the average Joe. I'll try to find it.
 
It’s funnny you mention an Ashley amp. I was setting up a room I was to work in. The owner had me get an Ashly amp for the speakers. I think it was 150w p/c. I of course didn’t want that Ashley amp but I tried it. Didn’t sound good.

I figured it probably just needed the brute force approach. I told the boss we needed another one. Got a second one and tried each in mono. It sounded much better but still not as open as I thought it would.

I told the boss we needed another pair. I bi amped them with each amp in bridge mono.

The speakers sounded good then. Those speakers were a later Dahlquist model. Then we got a pair of B&W 808’s and used the Ashley’s on those. Sounded good. The 808 are the only B&W’s I’ve ever really loved.
 
user 37518 said:
P.S. Being an audio engineer doesn't mean that much,

Lol. I believe you, but I'm both and average Joe and and audio engineer!

P.S. audio engineers probably have way more bias going into tests because they think they know what things are going to sound like.
 
Gold said:
It’s funnny you mention an Ashley amp. I was setting up a room I was to work in. The owner had me get an Ashly amp for the speakers. I think it was 150w p/c. I of course didn’t want that Ashley amp but I tried it. Didn’t sound good.

I figured it probably just needed the brute force approach. I told the boss we needed another one. Got a second one and tried each in mono. It sounded much better but still not as open as I thought it would.

I told the boss we needed another pair. I bi amped them with each amp in bridge mono.

The speakers sounded good then. Those speakers were a later Dahlquist model. Then we got a pair of B&W 808’s and used the Ashley’s on those. Sounded good. The 808 are the only B&W’s I’ve ever really loved.

Wow those 808's are serious! I've never heard them but would like too. I don't like what I've heard of the newer B&W's for mastering, way too bright. I now work on PMC MB2's. They were a little bright and it took some getting used to but now I love them.
The B&W 801's were my home deal a while back. I liked them a lot with that Hafler and the Hypex. The 801's had a weird passive circuit breaker thing. They had a bunch of diodes that would fill up a cap at a certain point and pop a relay. I took that out. I recapped all the caps in the crossover with film caps. I felt like they sounded better that way, but who knows!

I guess if you throw enough wattage at a speaker eventually you'll get there ;D
 
I always thought the high end of the 801’s and 802’s sounded weird. The only amp I’ve ever heard that could drive them where they sounded good was a giant Krell. Each one the size of o coffee table.
 
BTW the console I’m building has four identical stereo signal paths. Conducting listening tests was baked into the design. I can do long term listening tests which ABX can’t take into account.
 
Gold said:
BTW the console I’m building has four identical stereo signal paths. Conducting listening tests was baked into the design. I can do long term listening tests which ABX can’t take into account.

Yeah, theres a lot of stuff that people use to discredit ABX, stuff like "its different in the comfort of my own place", "you need to burn in the equipment", "I need to get used to the acoustic space", "you need to listen for a week to really hear it", etc., etc., etc.. of course there is always the good old "The ABX box makes it sound different"
 
user 37518 said:
There is a difference between art or the creative process, for example when doing a mix or producing a song, and another is the difference between what can be proven or not. In the medical industry they realized about this very early on, thats why they don't tell patients they are taking the placebo, that influences not only their actions but some of them actually feel better thinking they are taking the real thing, also, the research doctors don't know which patients are taking the placebo or the real drug because it also influences their findings or conclusions.

Thats the entire logic behind double blind testing, if professionals are so confident on what they can hear, then there shouldn't be a problem with double blind ABX test, the problem is that in the conditions I've been explaining on all my previous posts, people can't hear it, and thats why they try to disprove it with many (pointless) arguments.

You seem to be neglecting who has what vested interest. The drug industry certainly has a vested interest in their drugs working (and doctors to some extent as well), so the tests are a necessity.

With audio it's different,  some want differences,  others don't,  so who is doing the test and why are important considerations. They are usually poorly done often designed to produce a desired result.

So I prefer to do my own tests and make my own conclusions when possible. In my own experience I can't really speak to power amps,  but I've done a bunch of ABX on other things.  My predisposition is to get the best value,  to get the same for lower cost is a win. But it doesn't always work this way. Ild rather the cheap gear sounded just as good,  but it often doesn't.  Ild rather 96k didn't sound better in my setup,  but it doesn't.  Ild rather cheap beer and cheap wine tasted just as good, but it generally doesn't.

Another often neglected aspect of audio ABX is how draining critical listening can be.  I've had numerous experiences of doing nuanced tests where they started out 70+% correct and descended into coin flips.  After a long enough break the pattern re-emerged.
 
user 37518 said:
Yeah, theres a lot of stuff that people use to discredit ABX, stuff like "its different in the comfort of my own place", "you need to burn in the equipment", "I need to get used to the acoustic space", "you need to listen for a week to really hear it", etc., etc., etc.. of course there is always the good old "The ABX box makes it sound different"

I’m not sure who you are arguing with. I haven’t made the golden ears argument. I said I could design a double blind whoopee ha ha that the average Joe could tell the difference with speakers and power amps.

I haven’t discredited ABX. I have said that it is difficult and impractical. It is difficult to conduct fair listening tests and I have built a console that allows me to conduct better listening tests than anything but a double blind whoopee ha ha.
 
I've participated in listening studies, the really strict ones that require you to take an audiometry test before the test, all of them are double blinded, instead of relying only on "I know what I can hear", they would laugh at your face if you tell them that.

You said you could provide an easy test to listen for a difference, then you said you couldn't and were not interested, then you say again that you can design a double blind test that the average Joe could tell a difference, make up your mind.

You should publish your findings in a respected journal if you think its better, of course, you first have to convince them to let you publish such questionable results. Tell them your console is special  whoopee ha ha.
 
user 37518 said:
I've participated in listening studies, the really strict ones that require you to take an audiometry test before the test, all of them are double blinded, instead of relying only on "I know what I can hear", they would laugh at your face if you tell them that.

I know what proper double blind tests consist of. I don’t discount the AES test. If no one could hear a difference in that setup I don’t doubt it.

Extrapolating that one study applies to all amplifiers and speakers in all situations would be bad science.

I do A B tests all day every day for the last 30 years. Like I said I can conduct better listening tests than any studio I’ve ever seen with my setup.

You said you could provide an easy test to listen for a difference, then you said you couldn't and were not interested, then you say again that you can design a double blind test that the average Joe could tell a difference, make up your mind.

You are not paying attention. It’s the one and only test I proposed. To set it up as a double blind whoopee ha ha would be difficult and expensive.

You should publish your findings in a respected journal if you think its better, of course, you first have to convince them to let you publish such questionable results. Tell them your console is special  whoopee ha ha.

Huh? I haven’t conducted any studies. I told you my experience and said I could prove it in a double blind whoopee ha ha. I’ll bet you the double the cost of setting it up and publishing the results in a peer reviewed journal.
 
bluebird said:
Lol. I believe you, but I'm both and average Joe and and audio engineer!

P.S. audio engineers probably have way more bias going into tests because they think they know what things are going to sound like.
I consider myself an audio engineer, I learned decades ago that I could measure flaws that I couldn't hear, and could measure pretty much anything that I could hear, while I did have to roll some of my own test equipment back in the 70s.

I have built up several prototype circuits with great expectations about what they would "sound" like. Sometimes resulting in deep disappointment, but that is why we melt solder. 

I haven't designed an audio circuit "by ear" since my studio efx days (Loft delay/flanger) back in the 70s. I recall one listening session in a friends studio when I was comparing alternate circuits, not blinded. We took a break where some present dosed up with a cigarette, I had a cup of crappy vending machine coffee. When we picked up the listening test maybe 15 minutes later, the sound seemed like it was night and day different than just before taking the break. I am pretty sure the efx circuit didn't change by itself, but my hearing perception clearly did. We were playing back session tracks from tape, and that didn't change either.

===
I am pleased with the sound quality of my 6 channels of Hypex class D in my home theater, but apparently I am not as critical of a listener as some here.

===
I still try not to argue with people about what they say they can hear. I know how unreliable my ears were, and they surely have gotten worse over the decades since.

JR

PS: Another old story, back when Peavey was promoting it's solid state tube emulation (transtube) they set up a single blind A/B rig in our NAMM show sound booth. Only a tiny handful from the hundreds of dealers and music industry professionals could tell them apart. Of course the design engineer could tell them apart because he knew what to listen for.
 
Dear Mr. Science,

You said you wouldn’t believe me unless I conducted a double blind whoopee ha ha. I’m willing to put up double that. Pulling a number out of my a** I’d estimate $100K USD. I’m willing to put up $200K. Should be easy money for you.
 
Back
Top