Our concerns about numbers, harmonics, distortion, square waves, THD, measurements

[Lumba Ogir]

Hi,
this time it`s about the extremely popular but likewise extremely meaningless THD, intended as a figure of merit to prove the excellency of amplifiers beyond a reasonable doubt, provided by simulation software employing ideal parameter values and mathematical constants, being completely unable to analyze course of events. In this way, it never disappoints you, you can`t get anything but great results to be supremely proud of, although you might have minor difficulty explaining the mediocre sound. Of course, that figure in close vicinity of zero has nothing to do with reality, but don`t be particularly sad about this inconvenience, THD is just as meaningless when measured reliably with adequate equipment, because it does not show the harmonic content, which has overwhelming and decisive importance. For instance, the perceived difference between a class B and class A amp having the same THD is huge, that is in favor of class A and not even a drastically higher level there would change that impression.
Furthermore, the brain does not process sounds according to the commonly postulated linear scale, rather according to its own scale. Persisting in forcing that "linearity" on the ear is condemned to fail.

 

[destroyer X]

Post your ideas about.


Lumba is especially invited.

ear plus brain cannot process that information.... the thing works slowly because bio-chemical.... the square wave when reach ear plus brain looses the 90 degrées corners, because system cannot do that.... corners turns rouded, and what goes to brain is another thing very different from a square wave..something rounded enters.

So...when we use square waves and we make our efforts to make the square wave reproduced correctly in shape we are doing almost nothing...as brain will perceive another thing.

Numbers and THD are nice indication..but real world is filled with noises, reverberations, atenuations, distortions, harmonics and the real thing is not perfect...what means perfection?..... to do all those things?...... could be..but studios control all stuff and produce to you something very synthetic..something very filtered... a controlled product.

To study psycho acoustic is something we need to understand, a little bit more, such kind of reality...we can listen an old portable battery radio that cannot reproduce frequencies bellow 250 hertz (cutted, blocked internally, filtered, small condensers into stage input) and the speaker cannot go so deep... and they did that to avoid you to listen the mains hum when using old style AC to DC converters..so...this way....not beeing able to reproduce low frequencies, the mains frequency would not bother...... but people listen bass... this may be memory or musical experience...something our brain does completing the sound...so.... brain is always processing, adjusting the whole thing.

If you love someone and that woman call you by phone...and the environment is noisy.... transit, horns, people talking, wind, noises from everywhere...but you love that woman..and she call you.... could you perceive you listen only her voice and nothing more..the noises are filtered, deleted, erased from the system... only her voice is listened...why?... brain did that.

So... we use numbers to explain and evaluate...but the study of psycho acoustic will be more competent to help us to evaluate....

Are you sure humans dislike harmonics?.... and real sound has not distortions?... i am not sure....i have suspections the opposite...we love harmonics...harmonic is music..and distortion may be fine, depending what kind of distortions.

Our digital world is virtual...what means power measurement when you use simulator (almost nothing)...supply will drop voltage...real world is another thing...speaker system will react with EMF... well.... numbers and simulators are indications of possible maximum power...real world is different.

Distorted guitars are nice...why?

regards,

Carlos

.....................................................

 

[destroyer X]

People may say:

If ear plus brain cannot process square wave 90 degrées corners, so, may recriate this wave shape when receiving rounded corners.

Yes... it is possible...but what will happens with the real rounded corners ... the one is rounded into the original...will it be "squared" too?

Do the wave forms... the cicle has encoding..say... something alike tracking indicators..tracking numbers?

Will brain recognize the square wave that had rounded corners because bio chemical processing to restore the waveshape?

Maybe square goes to sinus.... but we can perceive clearly what is sinus and what is square wave.... you see... brain is much more clever than we are... we are "conscience"...there are much more to study.

Carlos

 

[destroyer X]

In many ways.... we...humans...use to search for hairs into bald head

You see.... down the sixties or seventies they have made the IFH institute rules about how to measure amplifier power.

You can see how wrong we can be.... and was followed and accepted by everybody.... factories started to make their specifications following IHF rules.

They said you could plug into power amplifiers a burst of signal many times bigger than the maximum allowable... this means hard distortion...the output was crazy...and they did that with only one channel beeing used..... very crazy that.... unreal.... as amplifiers use two channels simultaneously and we do not use burst of signal entering with so big levels, as we do not listen distortions.

And only one channel.... having time to the supply to gain voltage after the last burst of signal have entered...so.... measurement was many times bigger than the real thing....

And this was the rule...and nobody discussed there.

We are treating humans alike osciloscope...we are not osciloscope!

Carlos

 

[destroyer X]

In the seventies, while i was into the University, studying Psychology

I have found interesting things.

I have decided to measure brain waves and audiograms.... i have selected groups.

The rich and snobish girls that only talked about their travel to Europe... always informing about wealthy.... very futile people...and they are a group...only talking between them... a close group

The simple people, not snobish, or poor, the ones was thanking God they could study in so famous and expensive University...humble people..and they have their own group.

The depressive people, the ones had not self confidence.... the ones decided for psychology trying to discover their own personal problems.


I was in the second group....and measurements into laboratory show me the groups had consistence into measurements.... i had different measurements into each one of those groups.

Well.... after one year the laboratory was closed and i could not study the subject anymore...also i had not time because working into a Broadcasting Television Station....

Since that moment i perceive people were different related how their "system" works...how brain works.... and frequencies were the results.

I could not continue... found nothing to study about...and found no one more skilled to give me inputs.

I see we continue delirating.... very crazy thing to bother about 0.001% THD when speaker produces a mess much bigger than that.

I am searching for inputs about.... more than 39 years thinking about and i have discovered few thing but i am still very ignorant...... but also i can see there are many folks much more delirant than i am.... following the "stablishment"... the "status quo"... without think about.

Perception... psycho acoustic is the matter that interests me.... and i think is something that should be interesting for many folks too.

Carlos

 

[destroyer X]

Well... i have provocated a lot...now will be waiting someone with courage

to contribute....it is hard to say those things.

Imagine when someone said world was rounded.

When people said earth travell around the sun and that we were not the center of the universe.

The theories of Einstein.

Everything sligtly different is scandall.

I have discovered nothing.... those things were around us since the start of high fidelity concerns...people decided to study oscilograms....but we are not osciloscope folks!

Carlos

 

[CBS240]

IMHO, distortions in the time domain are much more audible than harmonic distortions. By this I mean the relationship of high frequencies wrt the lower ones. This is one reason why some of us don't like large electrolytic coupling caps.

Also IMD can be quite harsh.

 

[unclejed613]

one way of looking at it is this:

every waveform is a mixture of sine waves.

THD is only valid within the audible range and a bit beyond (the ear does process information up to about 50khz, even though you can't "hear" it, as it's processed as spatial information).

if you add harmonics to a pure sine wave, you've changed it's shape and harmonic content. you have changed it's sound. amplifiers shouldn't change the sound. unless you're a guitar player and you WANT the amplifier to change the sound

 

[Ragnwald]

Very nice reading Carlos and Lumba, i agree.
But we have to remember it´s a long chain of components from the microphone at the recording-movement till the sound comes out of your speakers at your home.
Every singel stage ad some sort of distortion, and i think it´s confirmd that 0,1% distortion is appreciable.
Good sound-engineering has benefits for the whole chain and in the end your experience in the favorite listening chair.

 

[jkeny]

Bud's whole EnaBl thread is pertinent to all this - here we have a treatment that seems to change the perceived sound from a driver (agreed by most who have heard it) and yet there is no significantly measurable parameter to explain this sonic change.

And the same old arguments flared up on that thread between those who maintain that measurements are God & those who don't. Some people call this the objectivists Vs the subjectivists but I prefer to think of the subjectivists as those willing to concede that science can't explain everything (YET)

Bud has an interesting theory about hearing on that thread, I can't remember what he calls it but it relates to our biological survival mechanisms and the fact that our ear/brain system has been trained to hear a discordant noise among a background of noise - this might be the noise of a snake in the grass. So if there is a discordant noise our attention is immediately drawn to this - the same happens in music & our mind is forced to focus on this discordance = listener fatigue? So identifying what is discordant is the secret, perhaps?

There is another thread on here about a guy who had tinnitus and his theory on hearing after he rid himself of tinnitus by retraining his brain's area of attention in the sound field

I agree with Carlos, the focus of attention should not be on simulators & THD numbers, etc but how the ear/brain truly works.

 

[jkeny]

This is much the same as the way the eyes are biologically adapted for survival - they are most adept at sensing movement as would be needed for sensing threats (snake in the grass example) or if hunting prey.

What they are not good at is staring at images that are stationary so when we read words on a page for instance our eyes make continuous micro movements to make those fixed, stationary letters appear to move slightly - all this slight micro movement leads to reading fatigue & tiredness of the eyes.

I wonder what we can learn from all this?

 

[Bob Cordell]

Hi,
this time it`s about the extremely popular but likewise extremely meaningless THD, intended as a figure of merit to prove the excellency of amplifiers beyond a reasonable doubt, provided by simulation software employing ideal parameter values and mathematical constants, being completely unable to analyze course of events. In this way, it never disappoints you, you can`t get anything but great results to be supremely proud of, although you might have minor difficulty explaining the mediocre sound. Of course, that figure in close vicinity of zero has nothing to do with reality, but don`t be particularly sad about this inconvenience, THD is just as meaningless when measured reliably with adequate equipment, because it does not show the harmonic content, which has overwhelming and decisive importance. For instance, the perceived difference between a class B and class A amp having the same THD is huge, that is in favor of class A and not even a drastically higher level there would change that impression.
Furthermore, the brain does not process sounds according to the commonly postulated linear scale, rather according to its own scale. Persisting in forcing that "linearity" on the ear is condemned to fail.

You have succinctly articulated all of the misconceptions about the value (or lack thereof) of THD measurements. At the same time you have shown your lack of understanding of the value of simulation.

Both THD and simulation are valuable in the right hands, when carried out with insight, care and diligence. Just because those results are imperfect and limited (which they are) does not mean that one like yourself should take an extreme position in throwing the baby out with the bathwater.

THD has gotten a bad rap by those who quote a single number, especially those who quote 1 kHz THD. That is about as valuable as testing the road-worthyness of a sports car by backing it out of your garage.

THD with full spectral analysis is, however, quite valuable. Similarly, viewing the waveshape of the THD residual is quite valuable.

Cheers,
Bob

 

[jkeny]

Both THD and simulation are valuable in the right hands, when carried out with insight, care and diligence. Just because those results are imperfect and limited (which they are)........

What would you say are the limits of these measurements, given the above?

THD with full spectral analysis is, however, quite valuable. Similarly, viewing the waveshape of the THD residual is quite valuable.

And again how much can you tell from this?

I'm not being contentious here, I'm just interested in the perceived limits of these methods. What important aspects of amplifier behaviour would you say is beyond the capability of these methods to measure?

 

[Bigun]

I have discovered through my interest in photography that no amount of measurement tells you which lens is best and that one man will disagree with another over which is best. But most people will appreciate a good photograph.


'not everything that can be counted counts, and not everything that counts can be counted' - Einstein

'And what is good, Phædrus,
And what is not good...
Need we ask anyone to tell us these things?' - Robert Pirsig

 

[ostripper]

By DX -The depressive people, the ones had not self confidence.... the ones decided for psychology trying to discover their own personal problems

That caught my eye.. would the psychological
attitude of the listener make his opinion subjective ??

As a "depressive person" , I'm not particurally impressed with
60 transistor PPM amps with thier .0001% specs. A good
15 device "schoolbook" amp can sound better if properly
compensated.

The best way to assess an amp is to "live with it" for a period
of time, your ears will either enjoy the experience or
say "gotta build a better one" .
Even designs that are
"golden" in simulation and measurement have been ultimately
rejected by the subconscious. So..why? I think cordell has
it right , the spectra of the distortion is more relevant than
"how much".

Output level is another psychoacoustical factor... I have had
designs that sounded good at low levels become undesirable
when driven hard even through they measured well at all
levels.I would rather have the .01% amp that sounds well,
as opposed to the ppm amp that is both harder to build
and offends my senses.
OS

 

[glt]

I have discovered through my interest in photography that no amount of measurement tells you which lens is best and that one man will disagree with another over which is best. But most people will appreciate a good photograph.

You really mean "no amount of measurement will tell which lens can take a good photograph" because taking a "good" photograph only depends on the photographer. But measurements such as resolution, barrel distortion, chromatic aberration and maximum aperture will definitely separate a good lens from an excellent lens.

But in engineering you have to measure or else there is no reference point upon which you can improve the product, or even do QA. At the moment there are no alternatives, unless you DIY.

 

[syn08]

You guys have no fun without an "objectivist" to argue with

Not me, this time

 

[Bigun]

You really mean "no amount of measurement will tell which lens can take a good photograph" because taking a "good" photograph only depends on the photographer. But measurements such as resolution, barrel distortion, chromatic aberration and maximum aperture will definitely separate a good lens from an excellent lens.

But in engineering you have to measure or else there is no reference point upon which you can improve the product, or even do QA. At the moment there are no alternatives, unless you DIY.

No that's not what I mean. In most cases good technical performance from a lens (amplifier) is a requirement to enter the game. But let's look at the list. How important is barrel distortion ? what level is unacceptable ? is it better to have higher resolution with worse barrel distortion ? the arguments would go around in circles. The results (photograph/sound) depend on the interplay between many specifications and also on human subjectivity.

What I really mean is that it is de-facto impossible to determine the quality of an amplifier or lens through measurement alone.

 

[bear]

Bob has the unique position of being correct and incorrect at the same time.

As he explained, THD alone is somewhat meaningless.

Looking at the residual distortion's spectra or just the shape of it can tell an experienced person something about how the unit is likely to sound.

I can personally tell you to a reasonably high degree of certainty how an amp will sound based on how the square wave looks on a scope. No kidding.

That's an important point, the ability via experience to correlate measurements of any sort, type or nature. This isn't automatic nor "scientific" nor "objective."

The other issue that comes into play is that this correlation depends on a number of non-objective issues and personal experiences that can not ever be transferred from one individual to another. It encompases everything from personal human perception (the ability to) through the quality of a given system that is doing the playback (notwithstanding any claims that it doesn't matter in some way or another).

In terms of design, measurements are a window on the invisible world of electrons and electronics, a very useful and valuable window.

The main question that it comes down to is what sorts of things or aspects of the way an amp or circuit operates, subtle or gross rise to some level of perceptable audibility, and under what circumstances.

This is nearly impossible to define in a quantified way since the threshold is not fixed in one place so to speak.

Someone mentioned that a non-linear transition from Class A to B in an AB or "B" amp is extremely audible. No doubt that this is a minute "glitch" and one that occurs near a zero crossing point. Audible? Not audible? Just an example.

You going to notice it playing back through all speakers? All rooms? Well that's the idea...

It comes down to what Count Basie once said "...if it sounds good, it is good..."

_-_-bear

 

[jkeny]

Hold on Bear,
You're saying on the one hand you can tell

to a reasonably high degree of certainty how an amp will sound based on how the square wave looks on a scope

But on the other hand you're saying that this is not based on any objective, or scientific basis and is non-transferable.

These two statements are contradictory and cancel one another out!

I would still like Bob to answer my question I asked above - what are the limits of these methods as far as telling how an amplifier will sound?