If you'd actually take a moment to stop talking in riddles, circles, and parables - maybe we could connect what's being discussed to something concrete. But essentially for the past few days, all you can say is :
"It's CMP"
How about some "maths" for a change? 🙄
We could address you as Mystic Michael if that would help motivate you into providing something that could be examined and tested. The constant riddles and talking in circles is no longer necessary. You now have a new title. Can we now have some real information?
Last edited:
Michael, just showing data without cross relating it to CMP is very different from just saying "it's because of CMP. I'm sure reference to any textbook will show how to explain things. If you are just trying to rename what has already been described otherwise, then it's just your terminology, not anything we can ever understand.
If you are able to show your own data, identify the problem of what you call "cmp", improve the situation, then you measure again to prove the situation is improved. Then we know that you really know what you are talking about.
If you are able to show your own data, identify the problem of what you call "cmp", improve the situation, then you measure again to prove the situation is improved. Then we know that you really know what you are talking about.
Last edited:
" Philosophical Michael" - *if* at all
😀
Its philosophy that embraces *how* to look at things – after all.
As said, CMP concept summarises and pin points a specific systemic pattern that seemingly was not recognized as such before.
Adding some "mystic touch" on well known mechanisms by putting up a single case of dubious significance on the other hand is more your terrain, I'd say
🙂
CMP is not exactly "just another re-labeling attempt"
What I've been stating with CMP concept is a point of view that is long overdue.
With respect to the current euphoria regarding impulse response shaping at will by all that fancy software and hardware now available for cheap – it tells us that "FR domain thinking" actually is a pretty limited concept by its own.
Michael
Let's see. How many time has it been stated that nonlinearity, while present, is not the dominant factor in loudspeakers.
You consistently repeat that nonlinearities are produced by resonance. Thus, it would follow that any resonant system must ultimately produce nonlinearities. The mathematics clearly contradicts that.
When you present you distortion measurements you look at the measurement as it defines the system rather than as what it is, an indication of the behavior at a specific operating point.
For example, the system equation is
O(t) = I(t) + 0.01 x I(t)^2
You look at the system when I(t) = 1. The output is O(t) = 1 + 0.01
The linear part = 1, the nonlinear part 0.01, -40dB.
Then you boost I(t) to 2 through EQ or what ever and observe that the nonlinear contribution is 0.04, (-28dB or an increase of 12dB) and make the claim that the EQ makes the system more nonlinear, when in fact, the system equation (and nonlinearity) has not changed in any way. You have consistently mistaken the measurement of the system state with the characteristic of the system.
You look at this nonlinear system and fail to realize that for inputs below 1.0, (or 0dB), the behavior is dominantly linear as the nonlinearity will never contribute anything greater than something -40dB or lower to the output. So we can say for this system that it behaves as a linear system to with in an error of no more than 1%. It is dominantly linear or, weakly nonlinear. The point of argument is not that nonlinearity is present or absent but that whether nonlinearity is significant over some operating range or not. The second argument is your continued referral to the observation of the change in magnitude of the nonlinear component in the output under different operating conditions as an indication that the system has become more nonlinear.
As for stored energy, I presume this is directed at Earl and I don't clearly understand where Earl is coming from. When developing the system equation there are several approaches. One is the common force balance, Newton's Law. The other is the energy method which states the the power delivered to a system is equal to the time derivative of the energy stored in the system plus the power dissipated.
Energy stored in a moving mass = mv^2/2, energy stored in a compressed spring is kx^2/2 , power dissipated in damper is bv^2
Thus,
Power in = d/dt(mv^2/2) + bv^2 + d(kx^2/2)/dt
= mv dv/dt + bv^2 + kx dx/dt
dv/dt = acceleration, a and dx/dt = velocity, v
Power in = mv a + bv^2 + kx v
Divide by v,
P/v = ma + bv + kx
Or,
F = ma + bv + kx.
Stored energy is always present in a dynamic system. It's just a matter of where it is stored and how much.
Frankly I would be please if the Moderators just cut off this discussion on linear vs nonlinear as it has degraded this thread and serves no useful purpose at this point.
If you like, move this aspect of the thead to a separate thread and close that thread. This is not what DIYA should be about.
All systems have a fundamental period after which their motion repeats. With nonlinearity, what happens is called "period doubling" - after a time the nonlinearities create signals at twice the period, and then twice that again, etc. etc. at some point the repetition period is longer than what you can measure (or are willing to wait for) and in essence the system is no longer periodic, it has become random and hence no longer predictable. The weaker the nonlinearity the longer this takes, but it will always happen. This is fundamental to the concept of chaos. Since all systems have some nonlinearity then all systems must go chaotic after some period of time (external control excluded).
I have spent my life studying vibrations and sound and in those studies energy is not a big part of this. I also have studied quantum mechanics and energy is the dominate poiint for view for this field. My issue is that "strored Energy" is being used here in a way that is neither clear or enlightening since there is, in essence no energy "stored" anywhere. In an simple single degree of freedom system like a mass and a spring we can put energy into the system, by displacing the mass for example, or hiting it with a hammer. This energy then goes into one or the other components, say the spring for a dsiplaced mass as potential energy and is then transfered into the mass as kinenetic energy and back to the spring as potential energy, etc. For a conservative system this total energy is always the same, but never stored in either one of the components. One could say that the btotal system energy for a conservative system is "stored", but the claim has been for the "masses stored energy", which is incorrect. Now in a non-conservative system, along the way some of this energy gets disapated in the damping (as heat, but actually the energy is still there!) and eventialy all of the intial mechanical energy is gone. There is "stored energy" only instantaneously and this was my point about having to consider what we mean by stored energy relative to a time scale. The energy is constantly moving about in an instantaneous sense and constantly being disappated in a longer term sense.
In quantum mechanics, systems can "store energy" for infinite periods of time and this concept has some meaning there. But for a SDOF resonate system? I don't see it.
I have never liked the term and this discussion is a clear reason why. It does nothing to make the discussions any clearer and, indeed, is used to hide falsehoods.
I'm glad you brought up doubling. Back in the day it wasn't hard to find a woofer that doubbled while at relatively low levels. Motors and suspensions were so poorly designed. It was even pretty common. But with today's designes, well when was the last time you worried about doubling when you selected drivers for your speakers? I remember Julian Hirsch of stereo review testing speaker for doubling at low frequency.
Wow, so much energy expended on such a simple concept. If 2nd order resonance was non-linear, how could the Fourier transform be used to describe it in both time and frequency?
What next, a heated debated that V <> IR?
What next, a heated debated that V <> IR?
I have data showing that EQ kills two birds with one stone, well, actually 3. Namely FR on axis, FR off axis, and improvement in CSD. So what do you think else could go bad? If it's something I can measure to prove you right, I certainly will post it.
I certainly would love to show some horn/waveguide data, but nothing yet has come up to my satisfaction. When I get another chance to do some measurements, I'd like to be able to show what happens to cone breakup. Surely there must be someone out there that had done basic research like this and reported it. I'd be surprised if I were the first.
Hello ,
Compact is good, I like the sketch. The centers of the drivers are close together. Go with the feeling and see how it works.
DT
All just for fun!
George,
The point Earl is making can be seen by example. On axis suppose you have a 3dB dip in the response at some frequency because of cancelation between the direct and cabinet edge diffracted sound. Off axis you might have a peak at the same frequency becaus eof the difference in the phase relationship. If you flatten the on axis dip, the off axis peak can get worse.
The point Earl is making can be seen by example. On axis suppose you have a 3dB dip in the response at some frequency because of cancelation between the direct and cabinet edge diffracted sound. Off axis you might have a peak at the same frequency becaus eof the difference in the phase relationship. If you flatten the on axis dip, the off axis peak can get worse.
Wrong put .
Looked at the system at a certain time slot, energy actually *is* stored in spring or mass.
That it „moves around“ along time line does not matter at all.
Besides some points along the discussion you made which I do not agree on, I appreciate that you brought it perfectly to the point with respect to the B***S term "stored energy" in general
😀
To expand a little bit more on the subject one could additionally state, that of course there is energy stored in an (ideal) mass-spring-system - be it in whichever component at whichever time slot – but that does not tell us anything - as with an omnipotent driving source we could make the system behave like whatever we wish.
The crucial point however is what can / cannot be accomplished with painfully limited resources like real world units
The really interesting point regarding these limitations is when there is *delay* involved.
Michael
So I can "store" something for only an infinitesimal amount of time? That seems inconsistant with the words definition and it makes ALL energy "stored".
It seems to me that EQ can also damp cone resonances (including breakup)as well. This is what I was trying to demonstrate.
If you hit the 10000th, I'll buy you dinner.😉
If you hit the 10000th, I'll buy you dinner.😉
If the resonace is a SDOF (single degree of freedom) then yes EQ can do that, but if it is distrbuted, like a cone resonance then it cannot because the response from this is not unifrom in space.
I'll be In Shanghai next week, so lets do that!
I think you are, with your discussion of “quantum mechanics” and all, perhaps reading too much into “stored energy” and as a result missing its “common language” meaning. The term is used to reflect something that happens over time, in a time frame (at least potentially) significant to human perception. If one hits a board with a hammer the energy of the blow is dissipated (internally damped) and radiated quickly . . . and we get the sound of, well, a hammer hitting a board. Deliver the same blow to a bell and little of the energy is damped, while most is “lost” by radiation over a perceptually significant period and we hear, well, the peal of a bell. “Stored energy” refers to that extension in time between the exciting impulse and the resultant radiation of that energy as sound. It’s real, it’s undeniable, it’s measurable with impulse response tests, and, in extreme cases at least, we can hear it. Maybe even in loudspeakers . . .
+1
There seems to be a (too common) tendency in speaker design to try to force reality to fit an oversimplified model . . . and treating cone breakup as a simple resonance (because that's how it looks at one point in space on a smoothed response curve) is a good example of that. It is, unfortunately, far from the only example. Simple models work very well for simple problems . . . we've got sub-woofers in free space down pretty well. Funny how that model fails to fully describe even that simple system when you put it in my living room.
- Status
- Not open for further replies.