Wolf, General Electric proved that that the human ears uses odd ordered harmonics to determine the volume of a sound back in the mid-60s. Its not really open to debate anymore- its more about- what to do about it?
IMO what their research indicates is that you really don't want to add any odd ordered harmonic distortion, even in vanishingly small amounts if you can at all help it. BTW this is why two amps can measure the same bandwidth on the bench, but one will be bright and the other not.
*Real* dynamics are indeed the difference between loud and soft, but in audio it usually is more about the distortion. When you get rid of it you find that the system can play any volume without sounding loud- its always relaxed, yet there is plenty of punch, enough to shake the walls.
Now there are ways that audio products take advantage of this psycho-acoustic phenomena. An example is the SET amplifier. They have high distortion at full power and unmeasurable distortion at low power. As the music plays louder, the transients are where the real power comes into play. At low power what distortion the amp makes is mostly lower orders (2nd harmonic) but on transients where more power is required the odd orders come into play. This results in the loudness cues being present on the transients (as far as the ear is concerned). This is why SETs are known for being a lot more dynamic than their small output power would seem to indicate. Effectively, the distortion is masquerading as dynamics.
If, on the same speaker you then substitute an amplifier that is lacking the odd ordered harmonics, you will find that you have to turn the volume up considerably higher to get the same apparent volume. But if you measure it with a sound pressure meter, then you will see what is happening.
I find that getting as much bandwidth as possible, along with the no odd ordered harmonics is the key to getting life-like presentations. This is very hard to do with transistors, as the non-linear aspects of the devices themselves contribute to odd-ordered harmonic distortion (and to be clear, we are talking about vanishingly small amounts such that it is difficult to measure, but our ears are more sensitive to these harmonics than just about anything else), making is very difficult to build a transistor amp that does not come off as bright or harsh (both are indicators of the presence of odd ordered harmonics).
This is not to say that you can't do it with transistors, its just that its a lot harder. I can count on one hand the number of transistor amps that do what tubes do easily in this regard, with fingers left over.
IOW this requires a re-think of how we do things. Take a look at this article:
http://www.atma-sphere.com/Resources/Paradigms_in_Amplifier_Design.php
The Power Paradigm is all about obeying human hearing rules, the Voltage Paradigm is about making the equipment look good on paper.
If anyone is having difficulty understanding how the ear uses odd ordered harmonics, there is a simple test I can outline that demonstrates it very effectively.
IMO what their research indicates is that you really don't want to add any odd ordered harmonic distortion, even in vanishingly small amounts if you can at all help it. BTW this is why two amps can measure the same bandwidth on the bench, but one will be bright and the other not.
*Real* dynamics are indeed the difference between loud and soft, but in audio it usually is more about the distortion. When you get rid of it you find that the system can play any volume without sounding loud- its always relaxed, yet there is plenty of punch, enough to shake the walls.
Now there are ways that audio products take advantage of this psycho-acoustic phenomena. An example is the SET amplifier. They have high distortion at full power and unmeasurable distortion at low power. As the music plays louder, the transients are where the real power comes into play. At low power what distortion the amp makes is mostly lower orders (2nd harmonic) but on transients where more power is required the odd orders come into play. This results in the loudness cues being present on the transients (as far as the ear is concerned). This is why SETs are known for being a lot more dynamic than their small output power would seem to indicate. Effectively, the distortion is masquerading as dynamics.
If, on the same speaker you then substitute an amplifier that is lacking the odd ordered harmonics, you will find that you have to turn the volume up considerably higher to get the same apparent volume. But if you measure it with a sound pressure meter, then you will see what is happening.
I find that getting as much bandwidth as possible, along with the no odd ordered harmonics is the key to getting life-like presentations. This is very hard to do with transistors, as the non-linear aspects of the devices themselves contribute to odd-ordered harmonic distortion (and to be clear, we are talking about vanishingly small amounts such that it is difficult to measure, but our ears are more sensitive to these harmonics than just about anything else), making is very difficult to build a transistor amp that does not come off as bright or harsh (both are indicators of the presence of odd ordered harmonics).
This is not to say that you can't do it with transistors, its just that its a lot harder. I can count on one hand the number of transistor amps that do what tubes do easily in this regard, with fingers left over.
IOW this requires a re-think of how we do things. Take a look at this article:
http://www.atma-sphere.com/Resources/Paradigms_in_Amplifier_Design.php
The Power Paradigm is all about obeying human hearing rules, the Voltage Paradigm is about making the equipment look good on paper.
If anyone is having difficulty understanding how the ear uses odd ordered harmonics, there is a simple test I can outline that demonstrates it very effectively.