The reason for my "pronouncement" ;-) about the 85% contribution of the amp/speaker/speakercable "group" as it affects system sonic character is as follows:
A. The amp is where the signal gets processed the most -- multiplied many many times its original strength and therefore subject to all manner of (possible) modification by the amplifier's electronics: harmonic and intermodulation distortion, variations in frequency distribution, timing changes, noise components and many others. If you ever heard the effects of a microphonic tube for instance, you have some idea what can happen on a less obvious level. All equally true for SS amps as well.
B. Loudspeakers have their own sonic signature(s) as we all know, but that's not the end of it. When an amplifier of one kind or another "looks into" the load characteristics presented by a given speaker (particularly the speaker's crossover network,) bad things can happen (or not) depending on the electrical compatibility of the two devices.
C. Ditto for the electrical characteristics of the speaker cable in use. Example: electrostats require low capacitance speaker cable, which is why it's advisable to stay away from high capacitance cable (almost all Cardas for instance) if you're driving electostats.
Then there's the issue of quality, no matter what kind of gear you want to use -- there are certainly some VERY quiet (and VERY expensive) tube preamps as there are SS preamps. My point was that using a great tube preamp (instead of a great SS preamp) will not get you "tube sound".
So how is it a tube amp produces "tube sound"? And why can't (most) SS amps produce "tube sound"? The answer is quite simple really. but first I need to define what, for me anyway, constitutes "tube sound". To me "tube sound" means that every molecule of signal detail is PRESERVED as the signal is amplified, and then it is sent on to the speakers "whole". In other words, to me it doesn't necessarily mean a "warm" sound, or a "mellow" sound. "Tube sound" is most noticable in an extremely lifelike midrange -- a quality that is less important to rockers for instance, than dynamic range, power, and tight bass.
To understand why tubes 'preserve' the signal better than transistors, you need remember just one thing: "switching speed." Power output devices (tubes OR transistors) are actually valves that normally operate in pairs, often referred to as "push-pull". One valve (tube or transistor) turns on and lets a powerful voltage through that represents a half-wave of the signal. Then it turns off and it's partner opens and lets the other half of the wave through.
OK, here it comes:
Tubes don't turn on/off instantly. They heat up, they cool down. And so there is some overlap between when one of the pair stops operating and the other one starts. There's a bit of -- let's call it lingering -- in handing off the signal; and so "the ball can never get dropped", so to speak. But if the "lingering" goes on too long, the waveform is distorted at the critical inflection point and you get muddy sound. (Some call it warm ;-)
Now transistors are just the opposite. They can turn on and off INSTANTLY. No lingering ;-) But what happens if when one valve turns off, the other one turns on just a nanosecond later? Sound (actually 'signal') was happening (or supposed to be happening) during that nonosecond. Where did it go? Well, it fell through the (nanosecond) crack! Lost! The result can be sound that feels 'hard' or 'etched', but in reality, the signal has simply lost some of it's original microdetail. It's a bit like when you throw away every hundreth pixel of a digital image, the picture seems to 'sharpen', but no longer looks completely "natural".
Very VERY expensive SS amps, like darTZeel, don't have cracks between their pairs of power output transistors ;-) However, it takes incredibly complex circuit design, and component matching, to reduce that nonosecond to a MILLIONTH of a nonosecond ;-)
.
A. The amp is where the signal gets processed the most -- multiplied many many times its original strength and therefore subject to all manner of (possible) modification by the amplifier's electronics: harmonic and intermodulation distortion, variations in frequency distribution, timing changes, noise components and many others. If you ever heard the effects of a microphonic tube for instance, you have some idea what can happen on a less obvious level. All equally true for SS amps as well.
B. Loudspeakers have their own sonic signature(s) as we all know, but that's not the end of it. When an amplifier of one kind or another "looks into" the load characteristics presented by a given speaker (particularly the speaker's crossover network,) bad things can happen (or not) depending on the electrical compatibility of the two devices.
C. Ditto for the electrical characteristics of the speaker cable in use. Example: electrostats require low capacitance speaker cable, which is why it's advisable to stay away from high capacitance cable (almost all Cardas for instance) if you're driving electostats.
Then there's the issue of quality, no matter what kind of gear you want to use -- there are certainly some VERY quiet (and VERY expensive) tube preamps as there are SS preamps. My point was that using a great tube preamp (instead of a great SS preamp) will not get you "tube sound".
So how is it a tube amp produces "tube sound"? And why can't (most) SS amps produce "tube sound"? The answer is quite simple really. but first I need to define what, for me anyway, constitutes "tube sound". To me "tube sound" means that every molecule of signal detail is PRESERVED as the signal is amplified, and then it is sent on to the speakers "whole". In other words, to me it doesn't necessarily mean a "warm" sound, or a "mellow" sound. "Tube sound" is most noticable in an extremely lifelike midrange -- a quality that is less important to rockers for instance, than dynamic range, power, and tight bass.
To understand why tubes 'preserve' the signal better than transistors, you need remember just one thing: "switching speed." Power output devices (tubes OR transistors) are actually valves that normally operate in pairs, often referred to as "push-pull". One valve (tube or transistor) turns on and lets a powerful voltage through that represents a half-wave of the signal. Then it turns off and it's partner opens and lets the other half of the wave through.
OK, here it comes:
Tubes don't turn on/off instantly. They heat up, they cool down. And so there is some overlap between when one of the pair stops operating and the other one starts. There's a bit of -- let's call it lingering -- in handing off the signal; and so "the ball can never get dropped", so to speak. But if the "lingering" goes on too long, the waveform is distorted at the critical inflection point and you get muddy sound. (Some call it warm ;-)
Now transistors are just the opposite. They can turn on and off INSTANTLY. No lingering ;-) But what happens if when one valve turns off, the other one turns on just a nanosecond later? Sound (actually 'signal') was happening (or supposed to be happening) during that nonosecond. Where did it go? Well, it fell through the (nanosecond) crack! Lost! The result can be sound that feels 'hard' or 'etched', but in reality, the signal has simply lost some of it's original microdetail. It's a bit like when you throw away every hundreth pixel of a digital image, the picture seems to 'sharpen', but no longer looks completely "natural".
Very VERY expensive SS amps, like darTZeel, don't have cracks between their pairs of power output transistors ;-) However, it takes incredibly complex circuit design, and component matching, to reduce that nonosecond to a MILLIONTH of a nonosecond ;-)
.
