The closest approach: what amplification?

If you are going to make progress with audio, the first thing you have to do it look at the way the human ear perceives sound- IOW understand the rules of human hearing.

Unfortunately while the bench specs we are so used to were being developed, research was going on at the same time that proved that much of the bench specs were/are meaningless, but the audio industry chose to ignore this research.

Its a common phenomena to ignore things that appear to be small errors, but if you study chaos theory you find that you do this at your own risk! It turns out that the way the human ear perceives volume is not by the actual sound pressure of the sound in question, but instead by the trace amounts of certain odd-ordered harmonics, the 5th, 7th and 9th to be exact.

Any enhancement of these harmonics is instantly heard, even if only hundredths or thousandths of a percent!!

Global negative feedback, which is found in most amplifiers, **enhances** these very harmonics while otherwise reducing distortion. As a result, it can be safely said that added loop feedback to an amplifier will violate one of the fundamental rules of human hearing.

However the presence of distortion will mask detail and cause harshness (even with even-ordered harmonics) so you have to prevent it. Without loop feedback, this is a bit chanllenging, but it can be done: class A operation, Triodes, simple circuitry, fully balanced operation (cancels distortion throughout the amp) and an avoidance of things like pentodes, transistors and transformers that are known to increase distortion excessively is one technique.

Of course, such an amplifier will have a higher 'output impedance' but it turns out that the effectiveness of things like damping factor in speakers is highly qualified in any event. All speakers, if overdamped, will exhibit loss of bass (up to 8db) and transient detail, and how much damping an individual speaker needs can vary over a range from 0.1:1(!) to 20:1. However, there are no known speakers that benefit from more than about 20:1; something to keep in mind!

The point is that if a speaker cannot be driven by an amplifier that is in fact designed to obey human hearing rules, its likely that there is also no way that speaker will ever be heard to sound like real music. That should stand to reason, but I am always amazed at how many people point to bench specs- even when their own experience is that those very same specs tell them nothing about the sound of the amplifier in question! That alone should tell you that bench specs are measuring the wrong things :)

Some references: http://en.wikipedia.org/wiki/Chaos_theory

"Critical Damping: Missing Link in Speaker Operation Parts 1 & 2 http://www.pearl-hifi.com/06_Lit_Archive/07_Misc_Downloads/Misc_Downloads.html

"IOW understand the rules of human hearing" is surely an overstatement. Hearing evolved along with the brain, and our survival and social needs for eons. We are a very long way from a complete understanding how it interprets information.

Atmasphere: I must be getting tired because I misread your opening paragraph.
Please ignore my response.
cheers

Atmasphere, i am not sold on the notion that audiophiles need to be good physicists or neural scientists to make progress with audio. as far as i am concerned, i much prefer making progress with audio by educating my ears to perceive dimensions about sound reproduction that one is not aware of unless he educates his hear. so, my approach is to educate my ear and then experiment with different systems, not studying the physics of sound or the neural bases of hearing.

Ggavetti, I completely agree with you. To be clear here though, its not the audiophiles that need to understand this stuff, its the **designers**!

IOW when you see designers building equipment that follows the human perceptual rules rather than bench specs, then meaningful audio progress will result.