http://users.ece.gatech.edu/~mleach/papers/TubeVsTrans.pdf

It involves a series of AB tests on experienced guitar players in terms of judging frequency response, ripple distortion, electrical impedance effects, and a number of other areas. It actually turns out according to this article the main reason that a tube amp sounds so much different from solid state is the different frequency response created from the speaker impedance load.

The idea of asking a bunch of pro guitar players to judge the effects of valve v. SS is about as likely to arrive at a useful conclusion IMO.

If I were going to build a guitar amp, I would absolutely enlist the help of some players so as to determine how to tune and adjust the response of the circuit. But not to analyze said circuit...
regards,
Douglas

I noticed the underdamped response curve right off the bat. That indicates a serious underdamping in the tube amp's output circuit, and probably represents a nearly worst-case situation. The transistor amp is pretty well damped, so its response curve is flat. But the tube amp circuit droops in the midbass and peaks in the bass.

If this paper were written to describe all tube amps, I would have objected to this as an oversimplification. Not all tube amps and loudspeakers will exhibit this anomaly. It's an obvious condition of high output impedance with a reactive loudspeaker load. It's a textbook case, really. This is what happens when you drive a reactive circuit with a constant current source. There are things that are done to address this in a high-fidelity tube sound system.

But this paper was written about guitar amps and not about hifi. It was written specifically about the Fender Twin Reverb amplifier. It's a nice amp and was popular with guitar players when this paper was written. I'd have to say that I agree with the author completely. It does exactly what he says it does. Not a bad thing, mind you. It's perfect for what it is used for.