Mean Thoughts on Measured and Perceived Performance.

Status
Not open for further replies.
Jump to Latest
#22
Lumba Ogir said:
In power amplifiers GNF introduces nasty distortions during the delay period and the more persistent recovery period exposing the amplifier for sharply increased susceptibility to overloading due to maintained open-loop gain, violent overshoots due to reduced bandwidth and slew-limiting, abrupt clipping due to constricted dynamic margins, furthermore phase intermodulation distortion (PIM).
Click to expand...

Hi

I just wanted to point out that PIM seems to be a function of the non-linear variables inside the real transistor model, and the open loop gain but not so related to closed loop GNF as you've stated. Now it is possible Mr. Cordell has no clue as to what he is talking about, but I highly doubt that.:whazzat: I agree it is an interesting subject as phase distortions may be more audible as coloration and loss of fidelity and should be discussed in more detail.
I try to look at this forum as a learning reference and there are some very interesting technical discussions here, as opposed to any other audio forum I've read. This is only because of the knowledgeable people here.

There is one major thing that separates us from apes, and that is not the ability to learn or to figure things out, but rather our natural inclination to teach each other, allowing us to build on developments from generations past. This forum is an outstanding example of this.😉
 
#23
Lumba Ogir said:

In reality everything happens in time.
Click to expand...

Not quite.

The human ear/human brain have two completely seperate perception and analysis pathways for sound (although the cochlea is used to sense both). One is based on fine frequency discrimination of about 3 to 10 cents that is relatively blind to time. The second is a time/energy pathway that is nearly blind to frequency. It detects the raw audio energy in a given time bin.

These two fundamentally different sensory inputs are reported to and analyzed by the brain using seperate neural pathways. They are reconciled (analyzed together) by the brain after independent analysis, then synchronized in time and backdated.

So, in the sense of perception forming reality, no, not everything happens in time. Especially given that the brain is backdating most of its analysis. You're actually 'hearing' things ten to a hundred milliseconds or so after you think you are!
 
#26
h_a said:
Fourier-theorem proves that every signal can be decomposed into a sum of harmonic waves. The Fourier-theorem is a scientific fact and contradicts your statement.
Click to expand...

But i posit that looking at just one of those sins misses a lot of information. The sheer number of of sins involved pushes the whole well into complexity theory.

As an analog look at the generator for the Mandlebrot set... it alone gives no idea of the incredible complexity and beauty that it can become.

Testing with single sin waves is a useful design tool, but a realistic test suite uses more complex stimulus.

Single number THD is a useless number... the spectrum that it is collapsed from is very useful.

dave
 
#27
xiphmont,
please do not take it as a brusque rejection. In fact, I find your information interesting.
Is it result of recent research? Do you have any explanation for the purpose of that dividing up?
Could you tell more about the range of frequency?
 
#28
Lumba Ogir said:
xiphmont,
please do not take it as a brusque rejection. In fact, I find your information interesting.
Is it result of recent research? Do you have any explanation for the purpose of that dividing up?
Could you tell more about the range of frequency?
Click to expand...

Recent research: not really. The oldest psychoacoustic research in this vein goes back nearly 100 years with Ehmer's papers. That's very early though. Anatomic/dissection work goes back farther, but things didn't really get moving until the 1950s when the empirical human-subject audio tests and physiological/neurological researchers started reading each others' papers (I'm guessing at this last bit, but suddenly both sides started moving forward rapidly and playing off each other). I could try to dig up photocopies of papers for references, unfortunately all of this material is still under jealously guarded journal copyrights. JASA already got mad at me once about that 🙂 If you want to find the actual papers, I recommend finding a big research university that has public library access... cheapest/most complete way to do it.

The purpose for two pathways seems pretty straightforward: fine frequency discrimination filters are very slow. Fast reaction (time) filters can't discriminate frequency much or at all. Combine the two with analysis to yield a system far more perceptively powerful than either alone. Researchers originally focused only on tome descrimination, believing that's all there was. They found the resonant hair structures in the cochlea, but were confused as resonant filters would have to be much faster than the reaction time results they were measuring. Then they found the time pathway.

Frequency discrimination, depending on individual, is usually limited to the classic 20Hz-15kHz range. It is common for young, healthy individuals to have frequency discrimination to 20kHz. It is uncommon, but not vanishingly rare, for genetic quirks to grant pure-tone discrimination as high as 100kHz. This quirk, for example, correlates strongly with people of Scandinavian origin who manifest environmental athsma.

The time filters have a wider input frequency range. Although most people can't hear pure tones > 15kHz, most subjects perceived higher frequency energy input to the time filter. Inpulses, noise and fast attacks are not perceived in frequency, but very high frequency components (>50kHz) do contribute to perception of these events. I don't have an upper figure; the paper I'm thinking of only tested to 50kHz.
 
#30
Looking a t your your firs t post in this thread Lumbar, I can only assume that you are the real creationist around here. Please, no more lectures about the significant improvement an amplifier design gets with zero feedback. I've heard enough amps with and without feedback (solid state) to know that generalizations like that are, well to put it frankly, creationist.
 
#31
Listen and all will be answered

Hi All.

I never thought that I would find myself throwing my hat into the ring for such debates.

Anyway, here goes:

I have always enjoyed good debates. Fosters great discussion and feeds the minds of those adventurous individuals.

But it always seems to go off on a tangent, and become just a little prickly when the subjective vs. objective debate flares up. Distill this thread down to the fundamentals and that’s all it is.

To me it seems very simple. Is it not all about the music? When all is said and done, and an amplifier has been subjected to every conceivable test known to man, and some black magic powder thrown in for good measure, we sit down and listen.

Sure, there are some who will print the specifications of their amplifier and hang them up, nicely framed and look at them all night.

Then there are others who just want to listen to the music.

Love him or loath him, Ivor Tiefenbrun (Linn Audio) said it best.
“If it sounds good it is good, if it sounds better it is better”

Seems pretty simple to me.

But then again, maybe I am the odd man out.

GV
 
#36
CBS240,
exactly, it`s about that line. Again, I impressed by your sharp observation pointing out a disgusting incorrect statement.
Our boundless effort to obtain knowledge is one of the human distinctive features.
Undoubtedly, this forum is a truly great and exiting place with a colorful multitude of opinions.
 
#37
xiphmont,
thank you very much for your clear explanation: separate sensory systems for time and frequency. While you possess excessive knowledge in this field, I am just trying to understand the physiology of hearing to some extent, however with growing admiration. The ear is a remarkable instrument but as the hearing evolved for specific purposes and environment, its rather peculiar characteristics (with reference to linear scales) should be taken into consideration in sound reproduction.
Well, hearing 100kHz tones certainly indicates serious asthma(!).
 
#40
xiphmont said:

The time filters have a wider input frequency range. Although most people can't hear pure tones > 15kHz, most subjects perceived higher frequency energy input to the time filter. Inpulses, noise and fast attacks are not perceived in frequency, but very high frequency components (>50kHz) do contribute to perception of these events. I don't have an upper figure; the paper I'm thinking of only tested to 50kHz.
Click to expand...

xiphmont, this is very interesting, do you have more information about this?

André
 
Status
Not open for further replies.
Top Bottom