Speaker impedance and amplifier distortion

Why is it not standard practice for amp manufacturers to publish 4ohm load distortion numbers? Obviously the numbers are likely to be worse due to the the higher demand on the power supply but if you’re going to sell your product as 4ohm stable you should publish the relevant info. Are the numbers so bad that no one wants to even touch them? Or am I way off base and the rule of thumb is that if an amplifier is considered stable at a lower load, it’s a given that distortion shouldn’t change?

Perhaps the specs aren’t a good indicator of how the amp will sound?

Not sure what that has to do with publishing the stats?

Well if you’re using a 4ohm speakers and shopping around for amps, wouldn’t you want to know how they compare? Otherwise, they might as well not publish any specs.

Well, I suppose I was more than implying the specs aren’t relevant. Even a very cheap, terribly designed amp will have better distortion specs than nearly every speaker ever designed and built

Those that makes amps know this and those the publish specs are pandering to those that care about such specs and possibly more than the sound attributes which are not measurable with current tools

Agreed. Specs are not a very good indicator of how an amp will sound so why go to the trouble of measuring and publishing them…

Good specs don’t indicate good sound, especially regarding amps, in my experience

Hmmm that’s a good point I hadn’t thought of…

I liken amps specs to using 0-60MPH specs to compare two different vehicles. The numbers really don’t tell you much about how each car drives or how it feels going from 0-60

My reply was directed at brett66.

Food for thought -


1 Like

I’m more interested in power amplifier output impedance and frequency domain transfer function into a real world loudspeaker load. Which always, with the possible exception of planar designs, has a significant complex component. That complex component can occasionally present to the amplifier a challenging, dynamically changing phase angle and very wide deviations from ‘nominal impedance’ in magnitude. This is why I’ve always valued JA’s published measurements in Stereophile. He fills that gap in manufacturer data sheets. His frequency response plots with a simulated speaker load are a godsend.

Aren’t the stats only valid for that specific simulation though?

Good points. I prefer to think of it as the reverse of the design approach it seems many take. Shoot for specific capabilities, then tweak for best sound. Look for good stats to narrow the field before listening.

Yes. I observe as a long time audiophile of something like 30 years the community continues to be highly polarized between the pure objectivists who take the view measurements tell it all and the pure subjectivists who take the view measurements tell us nothing.

I’m an engineer by trade. We accept in my business that no measurement is a perfect representation in every detail of reality. That does not mean we do no measurements or that measurements have no value.

I have straddled the fence from the beginning going back those 30 years. I equally weight objective measurements such as those published by JA and my subjective impressions from listening. Measurements in my humble opinion establish the basis for good design. Listening fine tunes that design.

So getting back to the use by JA of a simulated loudspeaker load, yes it is imperfect but I strongly believe it is a valuable datum in the overall decision process I make when choosing a component.

1 Like

So going by the logic that speakers’ own distortion will not allow you to hear amp distortion, it would follow that you shouldn’t be able to hear jitter either. Would you agree?

And if so, then would you also agree that a speaker’s own distortion would prevent you from hearing differences in audio or power cables?

Disclosure: I straddle the line between objectivist and subjectivist and therefore have no horse in this race. Just curious what people think.

Speaker inaccuracy is by far the largest source of playback distortion. But this inaccuracy is a fixed constant. Thus, one can perceive changes in the source, amplifiers, etc.

How can it be a fixed constant when the speaker is reactive?

Elk’s point, which I agree with, is the overall transfer function over frequency defined by the power amplifier and loudspeaker interaction is repeatable, not that it is flat. Which is correct. It doesn’t change from listen to listen. Other changes relative to that baseline can be discerned.

Exactly. But you explain it much better. :slight_smile:

1 Like

Got it, thanks. But I wonder how repeatable.