trendkillah nailed the basics. Here's a few subtleties:
The power supplies in most tube amps are 'soft', in that they are poorly regulated. To a typical engineer, this seems like a really shabby way to do things, but what happens, is that notes 'swell' as the initial attack drains off a lot of the energy stored in the filter caps, which is then built up again as the note sustains. Once the technogeeks got put right, (probably over several pints consumed in front of various bar bands), many transistor amps have been designed to reproduce this effect.
That sweet-sounding tube distortion [u]only[/u] happens when the [u]power[/u] stage of a tube amp is pushed into partial saturation. This causes the even-order harmonic content, and the compression, that characterise this sound. I've seen some interesting work based on FET output stages, but it isn't there, yet. The buzzy, overdive sound that a lot of players use, is generated in the preamp, or even between the guitar and the amp. Obviously, processing in this part of the signal chain does not interact with the speakers, missing an important part of 'the tube sound'. The plus side of that situation is the flexibility in volume level available to transistor amps that don't have to be cranked in order to find that 'sweet spot'. And preamp distortion sounds the same in headphones.
Transistor amps are ideal for accurately
reproducing a signal. A signal can be amplified without adding to, or taking away from, the original. Once a sound has been 'warmed up' by a tube stage in the preamp, a transistor amp will faithfully carry that warmth to the speakers. This is why modelling amps work as well as they do. What a transistor amp will
not do, is provide that warm, 'springy' touch that (so far, anyway) can only be coaxed out of a transformer-coupled tube power stage, reacting with a speaker cab.
Lordathestrings
Guitar Tricks Moderatorwww.GuitarTricks.com - Home of Online Guitar Lessons