Maybe I can put on my speaker designer hat and comment on Ralph aka Atma-Sphere's statement that you need to find out the intention of the speaker designer.
Let's say we have an "8-ohm" speaker whose impedance dips to 4 ohms in the upper bass region, and rises to 32 ohms at 3 kHz, where the crossover is (I'm picking exaggerated impedance swings here to make the math easier). If this speaker is voiced on a solid state amp, which puts out constant voltage, it will be voiced with the expectation that a 2.83 volt volt input into that 4 ohm dip will result in 2 watts into the speaker in that region, while a 2.83 volt input into that 32 ohm peak will result in 1/4 watt input in that region.
So, what happens when you hook this speaker up to a tube amp, which in many cases puts out approximately the same wattage into 4 ohms, 8 ohms, or 32 ohms? Well, now your speaker is getting only half as much wattage as it should get in that upper bass region, and it's getting 4 times as much wattage as it should get at 3 kHz! So we end up with a 3 dB dip in the upper bass, and a 6 dB peak at 3 kHz. Yuck! If we didn't know better, we'd conclude that tube amps generally suck.
But, what if this speaker was designed ("voiced") to be used with tube amps? Then it would sound fine on a tube amp, but with a solid state amp it would be 3 dB hot in the upper bass region and have a 6 dB dip at 3 kHz. Now, if we didn't know better, we'd conclude that solid state amps generally suck.
BUT if we knew the speaker designer's intention to begin with, we could have picked the right type of amp for the job.
Is there a way to build a speaker that works well regardless of the amplifier type? Yes, I think so. Briefly, the designer would shoot for as smooth an impedance curve as possible so that the tonal balance would stay pretty much the same when going from one amplifier type to another.
I've done some over-simplifying in this post, and left some issues unaddressed, but can come back and go into more detail if anyone wants.
Duke
