To upgrade or not to upgrade, that is the question.

[voodooless]

The notion that THD and SNR measured into a resistive load plus output impedance does not fully describe the performance of speaker amplifier is shared by many amplifier designers.

Let’s add a flat FR to the list…

It is possible that they are all either imagining things or simply lying to sell amps.

Why is it so hard te believe that the people selling amps lie? What about people selling € 50000 cables? What about the people selling those curious wooden contraptions that you put in your room (and no, those are not room treatment). What about the people selling stickers that you can basically put anywhere to “improve sound”?

It might also be that a lifetime spent building amplifiers has taught them some things.

So where is the science then? We’re not talking about magic here. It should be trivial to show the apartment problems certain amps have if they claim to sound differently. Then at least we'd have something to talk about. Now it's just: some designers have x years of experience, and they figured it out.

I’ve stayed longer on this forum than any other audio fora because believing in the patently ridiculous isn’t required.

So where is the line here? When is something patently ridiculous, and when is it not?

Almost everyone that’s been in this hobby a while (45 years) has had an amplification device that had good specs but after listening to it for months found something irritating about the sound. Replacing it with something else improved things. Yes, it might be explained by FR but not if a single damping factor number is all that matters. It might also be that the distortion profile was different driving a speaker than when heating a resistor. It might also be that the manufacturer was lying about the specs.

Or you didn’t do a controlled test…

The man asked for an opinion. Will a new low measured distortion Class D amp improve my system? My opinion is maybe yes, maybe no, might make it worse. Try one and see. If that’s not an allowable answer, ban me.

I’m all for trying, just add controls.

BTW, I just checked a Stereophile amplifier review. Atkinson uses the simulated speaker load for a frequency response test to show the FR effect of the output impedance. He uses resistors for distortion tests as does every other review I’ve seen.

Yes, they show that. And these differences can be audible in some cases. But again, there is no magic here. This is a function of output impedance. Would there be a reason not to buy the amp because it has a 0.1 dB deviation in FR somewhere? Would that justify the price difference? Also, correcting this with DSP should be trivial. If you have room correction, rerunning it with your new amp would fix it easily.

I'm not saying there can't be any audible differences. I will however state that in the vast majority of cases, they will be so small that you need controlled testing to do a good comparison. And in the vast majority of cases, any found difference will be explained by frequency response deviation, far beyond any other parameters that may differ.

 

[symphara]

Or you didn’t do a controlled test…

I love how this is thrown around, as if it's something you pull out of your pocket or buy from the corner shop for 2.99.

 

[voodooless]

I love how this is thrown around, as if it's something you pull out of your pocket or buy from the corner shop for 2.99.

You can't just ignore that this is an important aspect when wanting to do a fair comparison. Throwing around all kinds of potential reasons, while this one is the most common reason people hear a difference is just silly.

 

[symphara]

You can't just ignore that this is an important aspect when wanting to do a fair comparison. Throwing around all kinds of potential reasons, while this one is the most common reason people hear a difference is just silly.

Asking people to do controlled comparisons - the type with a signal switcher that I've never ever seen in commerce, a precise voltmeter that most people don't have, and a helper that knows what to do, so you can do it blind, is more of a bat to rethorically hit people over the head with than an argument.

It's of course super important and the final arbiter, at least for me.

Having said that, without having tested, just by listening, I can guess that between my Electrocompaniet and my 8500HA I couldn't tell the difference. However between the Primare A34.2 and the Electrocompaniet it was day and night. Similar power (lots) and THD specs (ok), very different sound. The D class was punchy with great, incredibly defined bass, but occasionally shrill, particularly with sax.

I'm sure it's not magic and there are measurable specs that manufacturers simply don't quote or even know.

 

[voodooless]

Asking people to do controlled comparisons - the type with a signal switcher that I've never ever seen in commerce, a precise voltmeter that most people don't have, and a helper that knows what to do, so you can do it blind, is more of a bat to rethorically hit people over the head with than an argument.

All I'm asking is to consider that for a decent comparison, this is essential. You can even start with a simple level matching to .1 dB or so. With a decent DDM, this can be done fairly easily.

It's of course super important and the final arbiter, at least for me.

Is it?

Having said that, without having tested, just by listening, I can guess that between my Electrocompaniet and my 8500HA I couldn't tell the difference. However between the Primare A34.2 and the Electrocompaniet it was day and night. Similar power (lots) and THD specs (ok), very different sound. The D class was punchy with great, incredibly defined bass, but occasionally shrill, particularly with sax.

I'm curious: what did the final arbiter say?

 

[symphara]

All I'm asking is to consider that for a decent comparison, this is essential. You can even start with a simple level matching to .1 dB or so. With a decent DDM, this can be done fairly easily.

Is it?

I'm curious: what did the final arbiter say?

Yes, it would be the final arbiter, as in it would settle the question for me, if I could do it. I don't have the equipment to do it.

 

[voodooless]

Yes, it would be the final arbiter, as in it would settle the question for me, if I could do it. I don't have the equipment to do it.

For the money that this equipment cost, you could have flown somebody half the way across the globe to help you set it up and do the testing And then have money to spare and buy just one of them.

 

[ahofer]

The man asked for an opinion. Will a new low measured distortion Class D amp improve my system? My opinion is maybe yes, maybe no, might make it worse. Try one and see. If that’s not an allowable answer, ban me.

Who said anything about banning? This is a forum, I disagreed. I also predicted your assertions would run into resistance, which they have.

And yes, Atkinson runs tests into load, that was my point. The source of sound difference is there in the measurements, not in some magical unmeasured behavior.

There have been many blind tests of amplifiers. One of the best was run by Stereo Review in 1987. They had a range of listeners compare Futterman OTLs to Levinson, Hafler, and a cheap pioneer. They couldn’t do it. So minor differences in measured FR and response to load are unlikely to be audible. There really isn’t evidence to the contrary. That doesn’t PROVE the point, but it would be nice if the believers in (and purveyors of) huge differences in similar-measuring amplifiers would step up and provide some scientific evidence. Until then, I go with the odds based on evidence at hand.

The idea that Class D will be audible per se also strikes me as unlikely.

 

[symphara]

For the money that this equipment cost, you could have flown somebody half the way across the globe to help you set it up and do the testing And then have money to spare and buy just one of them.

Well on the face of it your proposal leaves me with less gear

But you got the wrong idea. I got to choose between the Electrocompaniet and the Primare. I didn't buy the Primare because I didn't think it worked with my speakers. I got the Electrocompaniet instead.

The 8500 came years later. I think this AV receiver is as good for stereo, at least at my normal SPL.

In truth, I could get rid of the integrated amplifier and the streamer and just use the 8500. I don't do it because I wouldn't get much for them and I actually like them both.

 

[rwortman]

And yes, Atkinson runs tests into load, that was my point.

But he doesn’t run distortion tests into a simulated speaker load. That was my point.

 

[ahofer]

But he doesn’t run distortion tests into a simulated speaker load. That was my point.

Fair enough, but now we seem to be arguing about the definition of good-measuring or similar measuring: "What is a complete set of measurements?", as opposed to "is there something that can't be measured?"

 

[rwortman]

Fair enough, but now we seem to be arguing about the definition of good-measuring or similar measuring: "What is a complete set of measurements?", as opposed to "is there something that can't be measured?"

Correct. I never asserted that the difference was something impossible to measure. I said that I thought the measurements mentioned on the original post might be insufficient to characterize the sound of an amp and therefore it might sound different in either a good or bad way.

 

[ahofer]

Correct. I never asserted that the difference was something impossible to measure. I said that I thought the measurements mentioned on the original post might be insufficient to characterize the sound of an amp and therefore it might sound different in either a good or bad way.

OK. I'll reprint your comment, as that was not entirely clear.

There is more to the sound of an amplifier than those specifications you quoted. You really have to connect one to your speakers to know for sure. You might even find the class D amp an audible deterioration in sound quality.

It was the "know for sure" part.

 

[rwortman]

It was the "know for sure" part.

Well it is certainly a much easier thing to buy one and try it than to buy an analyzer to make measurements reviewers don‘t seem to be doing.

 

[voodooless]

Correct. I never asserted that the difference was something impossible to measure. I said that I thought the measurements mentioned on the original post might be insufficient to characterize the sound of an amp and therefore it might sound different in either a good or bad way.

But do we have any indication that measuring an otherwise good measuring amp with a speaker would exhibit a significantly different distortion pattern, than with a resistive dummy load? I'd like to see such an example.

 

[Spkrdctr]

I think the blind test part of any discussion can be made easier if the person tested switches the A/B products themsleves. Then after switching they listen with their eyes closed. That should get some seriously accurate results.

I offer that advice with humor.

 

[rwortman]

But do we have any indication that measuring an otherwise good measuring amp with a speaker would exhibit a significantly different distortion pattern, than with a resistive dummy load? I'd like to see such an example.

If I recall correctly, I asked about measuring amps while they are doing what they are designed to to, drive speakers, on another thread and Amir said it was annoying because of the sound and that the measurements were a mess that he didn’t know how to interpret.

 

[rwortman]

Many people have experienced an audio component that sounded good at first and after a long time began to grate. I don’t know what the explanation is for this. It does not lend itself to matched blind comparisons because no one listens at the same level for weeks or months and it’s unlikely they would listen for weeks or months and not peek at the equipment. I think it is possible that the we don’t listen the same way when we are conducting a test as we do when relaxing alone. I know that this is an easy “out” for the audiophile press to claim just about anything, but that doesn’t mean it isn’t possible. Mood effects all kinds of things.

Science is the search for new knowledge. Resolutely applying what science already knows is engineering. This forum, notwithstanding its name, is more about the latter than the former.

 

[voodooless]

If I recall correctly, I asked about measuring amps while they are doing what they are designed to to, drive speakers, on another thread and Amir said it was annoying because of the sound and that the measurements were a mess that he didn’t know how to interpret.

Thanks for that. Got any links?

 

[restorer-john]

@Ashley Salmond, your amplifier is good already, but we don't know it's distortion across the board and at elevated powers. Also, 120dB S/N with respect to 225W@8R is 42uV residual noise, so it's a quiet amp as it stands.

The ratings on the Purifi amplifiers would be better in some areas and likely poor or the same in others.

But you'd really need to listen to some other, comparable amplifiers and come to your own determination on whether you thing spending more money for a different amplifier is money well spent.