Inherent Design Question: Inherent sonic characteristics that cant be measured?

[Joel Wesseling]

Is there sonic characteristics generated from a driver that cant be quantified with todays best equipment and software?

 

[jcx]

are you talking Manhattan/Apollo/Human Genome Project

how much money and time do you have?

Ansys multiphysics sim running on a massive parallel compute farm?

some limits would be how closely you can measure your actual device material properties and dimensions
most speaker driver materials will vary in audibly important damping charateristics that won't be found in material handbooks, will be processing dependent to a large degree - can you test samples from the lots of materials used by the manufacturer for your specific drivers - or do you rely on NDT tests on the finished device?

at a more practical level Klippel is one source of info, test equipment: Introduction

 

[Joel Wesseling]

I know the general answer to my question and jcx points out things to help understand better..

What about thoughts on how much of the overall sonic characteristics of a great design is beyond the measured numbers .. I say a fair amount and this is the biggest difference between all well designed speakers.. Am I right?

 

[454Casull]

I'd say at this point basically everything we need to know can be measured. The only problem is, we don't know the causal relationships for everything yet.

 

[gootee]

I think we know all of the causal relationships, too, if we want to badly enough. But we draw the line, on both measurements and physics models, at some point of diminishing returns, i.e. "that's close enough".

 

[sdclc126]

I think if you can hear it -

you can measure it.

 

[planet10]

There are things we can't yet measure. And there is even more needs doing with correlating what is measured to what we hear.

dave

 

[Jay1111]

I dont really think the question should be is there something missing from the measurements. What needs to be resolved is, what order(s) of distortion, at what relative level(s), and at which frequency(s) is/are pleasing/displeasing.

 

[ScottG]

What about thoughts on how much of the overall sonic characteristics of a great design is beyond the measured numbers .. I say a fair amount and this is the biggest difference between all well designed speakers.. Am I right?

I think for what we currently know: we can pretty much measure *most* of what does (or does not) make fore a better measured loudspeaker (..a more linear loudspeaker that is).

There are two over-riding issues however that substantially "muddy the waters":

1. Just because you can measure doesn't mean you can correlate that measurement with an effect. Most mechanical effects - sure, but most subjective responses.. not so much (at least not currently). The issue I brought up on "depth" perspective (in the Orion/Behringer thread), seems to be a good example. People know what "depth" is - it's self-explanatory. IF system "A" provides enough of a difference from "B" in the portrayal of depth - then it starts becoming an obvious effect, but point to that on any graphical display of Impulse Response? Not really.

2. It's particularly bad when an objectively better design may well be incorrect for subjectively better stereo reproduction. Stereo is just plain *flawed* - designing the "best loudspeakers ever" may still result in overall mediocrity for any given listener. Honestly, because of this fact alone I don't think anyone will *ever* create a "perfect" stereo loudspeaker system..

Still, both aspects 1 & 2 are interesting impediments to attempt to overcome - and it tends to make the hobby that much more interesting. Some tend to scoff: "there's no *magic*!". But when no one can sufficiently provide an objective answer to an effect, then yeah - it's pretty much "magic" until it can be.

 

[picowallspeaker]

Stereo is just plain *flawed* - designing the "best loudspeakers ever" may still result in overall mediocrity for any given listener. Honestly, because of this fact alone I don't think anyone will *ever* create a "perfect" stereo loudspeaker system..

Oh yes, also a human is flawed
It has been misunderstood; recordings were made to make it eternal !
as we need to reproduce an event that happened in another time, in another place. So a recording has to bring to our ears what has happened then, which is a sequence in time and space...in a relatively very slow motion.
The cracks in the system allow so many variants that may corrupt this slow motion. Indeed what we hear is the processment of it all, nothing else.
So what we call stereo is just a double sequencing of a reality captured.
Making it synchronized to our expectations....so also the human being with its culture, emotivity or just brain subconcious activity in differentiating the arrival time and then elaborate, resulting ever time as a new event....like reading a book twice, or watching a landscape, but there's a little conceptual difference : you are hearing to something happened in another time, but that thing is happening again at your place

 

[davidsrsb]

I am puzzled at reviews that praise HEIGHT in the stereo image. There is no way two ideal speakers can give height clues that are not there in the original recording (floor and ceiling echoes maybe)

 

[454Casull]

I dont really think the question should be is there something missing from the measurements. What needs to be resolved is, what order(s) of distortion, at what relative level(s), and at which frequency(s) is/are pleasing/displeasing.

Only a few of the things which make a speaker sound pleasing and/or accurate.

 

[counter culture]

Is there sonic characteristics generated from a driver that cant be quantified with todays best equipment and software?

What would constitute proof one way or the other? What evidence would convince you that there are unmeasurable characteristics, or that vice versa, all characteristics are measurable?

I see you say you knew the answer already.

1. What was it that convinced you?

2. Why bother asking?

 

[RayCtech]

Is there sonic characteristics generated from a driver that cant be quantified with todays best equipment and software?

No - measurements cannot tell or quantify the sonic performance.

 

[Joel Wesseling]

My answer is NO.. Todays measurements dont tell the whole story of the sonic characteristics...Some Audible Vibration patterns are to complex to measure in its entirety with todays software and equipment..Those patterns impart a great deal on the sonic characteristics

Well, it all started with an engineer who is very knowlegeable in this area of design and measurement.. He states its all numbers and a great design is simply improving the numbers..

Okay as I go over it, He does state The difference between a good design and and a great design is Diffraction and Compression and that its hard to measure but its the only thing standing in the way between a good and great sounding speaker..

So Enlighten me with your thoughts and expertise..

Lets imagine Diffraction and compression are perfect in 2 speakers each with different driver. Do we still have a considerable sonic difference between the two? will we all hear them as 2 very different sounding speakers?

Just trying to learn

 

[jcx]

sonics = sound, physics - perception is another field entirely

if you know how all surfaces move its only a question of computer horsepower to get the sound field in a defined/known environment

several laser tools are availble today

Tools of the trade - laser measurements | Bowers & Wilkins | B&W speakers


microphone arrays, mechanical scanning could be correlated with radiation surface measurements, room multipath/modes/box diffraction...

again the physics isn't hard by PhD/top National Lab standards - the engineering, equipment cost and analysis time are expensive - really only a question of how far can you afford to go


I'm suprised that no one has tossed in an appeal Chaos yet - possible in driven mechancial resonace? - maybe - but sucessful speakers are fairly heavily damped
and proving Chaos in a physical system isn't trivial - it isn't the same as isufficently detailed modeling Chaos (Stanford Encyclopedia of Philosophy)

 

[RayCtech]

if you know how all surfaces move its only a question of computer horsepower to get the sound field in a defined/known environment

If it was so simple to design the perfect driver..

In my early stages of development I used laser and the other tools.
Without I would not have been able to "see" and rectify the flaws.
When this phase was finished and surround, cone, dust cap and voice coil was performing as wanted I had still not "listened" to the driver.

I now had 10 slightly different surround / cone assemblies that passed the above mentioned tests - 100 pcs of each variant for production variation control. 4 different voice coils. 4 different magnet systems (ferrite, alnico, two different neodym)...

It was at this stage when I tested out the 160 combinations I understood that traditional measurement methods do not tell anything about the sonics.

Later when I tested AMT prototypes I was able to verify that I could deform the diaphragm (with a heat gun) to the point where it was completely defect and no output at all.

It was very easy to listen and hear the drivers sound worse and worse - at some point there was more noise / distortion than sound.
However the measurements of harmonic distortion told that the driver was performing very good even at this point.....

 

[wintermute]

Personally I don't think it's a case of not being measurable, but more one of not knowing how to interpret or correlate the measurements to the subjective impressions.

Our brain is a supercomputer, and is programmed to pick up differences in sounds for many reasons. Even with enough computing power, we would still need to understand what it is that our brains are doing to be able to have a chance of developing visualisations of measurements that we could reliably point to and say this is a predictor of say phantom image quality.

Where it gets really messy is when we take into account the tricks that our brain can play too. Then it gets really murky.

I've mentioned before that I think it would be great to do controlled experiments with both sighted and blind listening where the subjects brain waves were monitored. I think this could be helpfull in improving the understanding between measurements and listener impressions. Of course interpreting differences in brain waves and what that means is another can of worms as well

Tony.

 

[gainphile]

Can we measure PRAT ?

Let's say "My speaker's PRAT score is 90" hence it's superior to your speaker's PRAT 75.

 

[jcx]

what's with the "standard measurements" strawman - I thought the thread title didn't preclude any tech, level of effort, expertise

and again the conflation of "sonics" with human perception - I claim sonics should be the science of sound waves == physics


it should too obvious to anyone here that there's no complete understanding of percption, interpretation of sound fields by diverse ears and brains - not worthy of yet another thread that can't possibly go anywhere