Are there any test or measurements that will predict which amplifiers will give best results with a particular speaker?
Picking the "best" amp for a particular speaker from measurements alone is probably unrealistic, but you can pick a best amplifier
type. And by "type" in this context I mean, constant-voltage (typical of solid state) or constant-power (typical of tubes) or constant-current (rare but Nelson Pass makes some).
Most loudspeakers are designed for constant-voltage amps, wherein the amp puts out increase wattage into the impedance dips, and decreased wattage into the impedance peaks. Some loudspeakers are designed for constant-power amplifiers, wherein the amp puts out approximately the same wattage regardless of the impedance curve. A few loudspeakers are designed for constant-current amps, wherein the amp puts out increased wattage into the impedance peaks and decreased wattage into the impedance dips, essentially the exact opposite of a constant-voltage amp.
None of these behaviors are necessarily "ideal", but the default is to design speakers for constant-voltage amplifiers, since this behavior is typical of solid state amps, which are
by far the most widely used. There are arguments for other amplifier types which imo are valid but not the focus of this post.
Why do some well known brands of audiophile speakers recommend the use of tube amps almost exclusively (Zu Audio comes to mind)?
This won't be the whole story, but it will be part of it.
Loudspeakers have one or more impedance peaks in the bass region, and a constant-voltage (typical solid state) amplifier delivers reduced power into these peaks. BUT a tube amp will deliver approximately the same amount of power into these peaks as into the rest of the spectrum, resulting in a
potential "free lunch" of deeper and/or more powerful bass.
So if we wanted to take advantage of that, or other, characteristics of tube amps, we would design the speaker to behave well when driven by a (constant-power) tube amp. The result might not behave well when driven by a (constant-voltage) solid state amp.
Let's look at an example of measurements of a speaker designed specifically to work well with tube amps: The Audio Note AN-E. Here is the frequency response curve that John Atkinson of Stereophile measured, and
note that he used a solid-state amp:
Looks pretty awful, doesn't it?
But remember this speaker was designed to work best with tube amps, which deliver approximately the same power across the impedance curve. Let's have a look at that impedance curve:
So a solid state amp will be delivering
less power than the speaker designer intended into the impedance peaks, and
more power than the speaker designer intended into the impedance dips. You might find this interesting, because we are going to uncover competence on the part of designer Peter Qvortrup that is not obvious from the measurements.
Let's start in the treble and work our way down, noting how a tube amp's behavior into that impedance curve would change the speaker's frequency response.
At 10 kHz the speaker has a frequency response peak with the solid state amp, but the impedance curve has a dip in that region, so a tube amp would be putting out
less power (relative to a solid state amp) in that region, and that peak would come down.
Just north of 1 kHz the speaker has a dip with the solid state amp, but the impedance curve has a peak in that region, so a tube amp would be putting out
more power and that dip would be filled in.
Just north of 200 Hz the speaker has a peak with the solid state amp, but the impedance curve has a dip in that region, so a tube amp would be putting out
less power and that peak would come down.
Centered on about 50-Hz ballpark the speaker's bass response has a broad "saddle" with the solid state amp, but the impedance curve has a peak in that region, so a tube amp would be putting out
more power and that saddle would be filled in. (This is the "free lunch" that I mentioned earlier.)
At 30 Hz the speaker has a small peak with the solid state amp, but the impedance curve has a dip in that region, so a tube amp would be putting out
less power and that peak would come down.
So as I hope you can see, there would be a welcome synergy between this speaker and a tube amp. I'm not saying the net result would be ASR-approved flatness, but it would be a significant improvement over the results with the solid state amp.
(In the absence of such detailed information about a speaker, follow the manufacturer's recommendation for amplifier types. If I was tasked with pairing amps with a specific speaker, I would cheat and research which amps the speaker manufacturer had chosen to use at audio shows, and those amps would be the ones on my short list.)
[Off-topic:] One final comment about this design: The (on-axis) frequency response curve we have predicted by assuming a tube amp instead of a solid-state amp would still be recessed in the 3-8 kHz ballpark. But this is at the bottom end of the tweeter's range, where there is excess off-axis energy, so we would WANT the on-axis energy to dip a bit in this region for the sake of good in-room frequency response. Here you can see that off-axis energy flare in the 3-8 kHz region:
Many people have looked at Stereophile's frequency response measurements of the Audio Note AN-E and concluded that the design is fatally flawed, but if we factor in the recommended amplification type and the off-axis response, the picture changes. In addition, the low frequency response is deliberately gently downward-sloping across the bass region to take advantage of boundary reinforcement from nearby walls. I have no affiliation with Audio Note.
