or Connect
AVS › AVS Forum › Display Devices › Ultra Hi-End HT Gear ($20,000+) › Digital Active Speaker Thread.....
New Posts  All Forums:Forum Nav:

Digital Active Speaker Thread..... - Page 11  

post #301 of 481
Thread Starter 
Quote:
Originally Posted by faberryman View Post

This is a classic example of the logical fallacy of false dilemma which permeates your thinking, and leads you to hyperbolic positions and untenable conclusions. Contrary to your assertion, speaker design can be both a math equation and a real world problem requiring human problem solving skills.

You were the one that said that designing a speaker is just a math equation, so I'm asking for the equation as maybe I'd like to build a speaker. Otherwise, it's all about human 'fuzzy logic' decisions/tradeoffs and the math is just a tool, not the problem itself.
post #302 of 481
Thread Starter 
Quote:
Originally Posted by faberryman View Post

So the NAD 9060 blows the Krell FPB out of the water, but only if you also buy a DEQX and then go home and take your speakers apart, remove the crossovers, wire in the DEXQ, learn how to use it, buy some test equipment and learn how to use it, and futz around hoping to find some settings which make your disassembled speakers both measure better and sound better than what you remember (if you can) your speakers originally sounded like. But now we are not really comparing amplifiers any more are we? It's no wonder more people don't do as you suggest. What are they thinking? How many customers have you convinced that this is a good alternative?

Or you could just use a $300 Behringer with the NAD and still be well under the price of an uber amp for higher performance, assuming there is no driver compensation. DEQX is just handier because it does the driver compensation and essentially optimizes driver behavior (well, corrects for it) before the crossover is engaged.
post #303 of 481
Thread Starter 
Quote:
Originally Posted by FrantzM View Post

John

The Mathematics, DeQX are , were said by all, to be tools.. Can you show me where anyone said that Mathematics were sufficient to build a speaker? Come On!!!!

Did you miss where it was said that it is simply an equation? That if you 'cheat' by removing some of the problems, that it throws the equation off and you end up with a poorer result?
Quote:



What we keep on repeating and that you childishly refuse to accept or try to distort is that a speaker can be modelized through mathematics.. The behavior of the driver can be modelized... What do you think the DeQX do? How do you think it corrects for driver aberrations? What do you think any digital crossover does? What do you think a crossover slope is? 48 dB per octave is a mathematical representation of the slope (another mathematical concept) of a crossover Isn't that maths?

Sure, but the math is a tool. The process itself is not an equation. It's like you saying that driving to work is an equation. And that if you put anti-lock brakes on your car, that it is more dangerous because it takes away one of the problems with driving.
post #304 of 481
Quote:
Originally Posted by Alimentall View Post

Or you could just use a $300 Behringer with the NAD and still be well under the price of an uber amp for higher performance, assuming there is no driver compensation. DEQX is just handier because it does the driver compensation and essentially optimizes driver behavior (well, corrects for it) before the crossover is engaged.

Of course, you still have to take the NAD and the Behringer home and take your speakers apart, remove or bypass the crossovers, wire in the Berhringer, learn how to use it, buy some test equipment and learn how to use it, and futz around hoping to find some settings which make your disassembled speakers both measure better and sound better than what you remember (if you can) your speakers originally sounded like. But then we aren't really comparing amplifiers, are we?
post #305 of 481
Quote:
Originally Posted by FrantzM View Post

What do you think any digital crossover does? What do you think a crossover slope is? 48 dB per octave is a mathematical representation of the slope (another mathematical concept) of a crossover Isn't that maths?

This is a great point. We often discuss a digital crossover as if it's the same as implementing an analog crossover. It's actually a pair of complimentary filters which approximates the performance of an nth order crossover, where n == an integer value of 6dB / octave.

So we would say that a 48dB digital crossover approximates the performance of an 8th order analog crossover. It's not perfect, but it's damn close.
post #306 of 481
Thread Starter 
Quote:
Originally Posted by faberryman View Post

But then we aren't really comparing amplifiers, are we?

Nope, we're comparing one way of doing it versus another. Context.
post #307 of 481
Thread Starter 
Quote:
Originally Posted by John Kotches View Post

This is a great point. We often discuss a digital crossover as if it's the same as implementing an analog crossover. It's actually a pair of complimentary filters which approximates the performance of an nth order crossover, where n == an integer value of 6dB / octave.

So we would say that a 48dB digital crossover approximates the performance of an 8th order analog crossover. It's not perfect, but it's damn close.

Sure, but even an analog crossover is approximating a theoretical crossover because the parts aren't perfect and the impedances aren't constant. The digital crossover will be closer to theoretically perfect.
post #308 of 481
Quote:
Originally Posted by Dennis Erskine View Post

I disagree here. Neither Physics nor Mathematics is magic...but Physics and Mathematics are magical in what they reveal to us.

Our current physics will do until the next set of even better theories surpass them. It took about 250 years to rewrite Newton with General Relativity. How long before General Relativity is rewritten (as an example).

Pardon me being slightly tongue cheek here. Physics (as you know very very well) is ever evolving
post #309 of 481
Quote:
Originally Posted by Alimentall View Post

Nope, we're comparing one way of doing it versus another. Context.

The question was a rhetorical one so that you would reflect on what the implied answer to the question must be. You correctly identified the answer as "Nope." So the $1500 six channel amp doesn't really blow the $20,000 amp out of the water. You haven't tried using the $20,000 amp in lieu of $1500 amp and made measurements to determine which is the better amp.
post #310 of 481
Quote:
Originally Posted by Alimentall View Post

Sure, but even an analog crossover is approximating a theoretical crossover because the parts aren't perfect and the impedances aren't constant. The digital crossover will be closer to theoretically perfect.

Maybe yes, maybe no.

It comes down to the implementation and whether or not you can live with the attendant artifacts. Some of them are very very very low (ie the 22bit range).
post #311 of 481
John

I rarely lose my temper on any thread but seeing your BS and steadfast refusal to concede on even the smallest point which is by the way a fact I am off your thread> You can BS all you want and leave yourself the impression of having won but you and everybody's in this forum know you are BSing big time....The facts are there and no amount of distortion spin and plain BS will not change it I am off do as you please with your thread... I have rarely seen such an immature display of intellectual dishonesty...
post #312 of 481
Thread Starter 
Quote:
Originally Posted by John Kotches View Post

Maybe yes, maybe no.

It comes down to the implementation and whether or not you can live with the attendant artifacts. Some of them are very very very low (ie the 22bit range).

Well, true, I was assuming bit depth at least equal to or better than the incoming signal. I still hear more artifacts in a passive crossover that I refer to as a type of signal compression that I don't hear on Meridians for instance.

BTW, I know you're on the 'ins' with Meridian, but have they told you they use 48dB/octave crossovers in the DSP7200? I can't find a reference to it either way. I am impressed by them, nice leap over the older models. I wish they weren't so expensive though.........
post #313 of 481
Thread Starter 
Quote:
Originally Posted by FrantzM View Post

I rarely lose my temper on any thread but seeing your BS and steadfast refusal to concede on even the smallest point which is by the way a fact I am off your thread> You can BS all you want and leave yourself the impression of having won but you and everybody's in this forum know you are BSing big time....The facts are there and no amount of distortion spin and plain BS will not change it I am off do as you please with your thread... I have rarely seen such an immature display of intellectual dishonesty...

Could you please explain what I am BSing about? And what point I am supposed to concede? I mean, I don't mind conceding if I'm wrong. You can not say that being able to correct, for instance, time in a speaker design electrically is going to cause some bad mojo because 'mother nature' can't stand having, what was it, an 'under determined equation' (?). That's why I say a speaker design ≠ an equation. Not without being overly simplistic. DSP is not a 'cheat'. It's an advancement. It does not through nature out of balance or equations out of whack. It solves problems. Elegantly. Leaving speaker designers to focus on other things that DSP can't solve. I'm sorry if you can't follow the logic on this, but there's no need to be angry, nor to lob insults.
post #314 of 481
Quote:
Originally Posted by Alimentall View Post

Well, true, I was assuming bit depth at least equal to or better than the incoming signal. I still hear more artifacts in a passive crossover that I refer to as a type of signal compression that I don't hear on Meridians for instance.

BTW, I know you're on the 'ins' with Meridian, but have they told you they use 48dB/octave crossovers in the DSP7200? I can't find a reference to it either way. I am impressed by them, nice leap over the older models. I wish they weren't so expensive though.........

It's probably similar to the 8Ks.

One woofer full to the midrange, one with an analog lowpass at 400Hz or so. The 8000 cascades even more, with one lowpassed at 400Hz, one lowpassed about 800Hz (per side). The analog lowpass could be as simple as an appropriately valued capacitor.

Then 48dB/octave woofer to mid. 48 dB/octave mid to tweeter. Complimentary filter pairs here of course.
post #315 of 481
Quote:
Originally Posted by Alimentall View Post

You were the one that said that designing a speaker is just a math equation, so I'm asking for the equation as maybe I'd like to build a speaker. Otherwise, it's all about human 'fuzzy logic' decisions/tradeoffs and the math is just a tool, not the problem itself.

John,

It's easy!!! For all the drivers, you write Newton's 2nd Law of Motion: F = ma.
That is Force = mass times acceleration. For the linear motor in the driver; you
need Faraday's Law. For the air in the box and surrounding the speaker; you need
the wave equation.

It's not a single equation; the equations are coupled. For example, the position of the
cone of the driver becomes a boundary condition for the wave equation describing the
motion of the air external to the driver.

Again, more complex systems than loudspeakers are modeled mathematically. Boeing
doesn't do much wind tunnel testing these days; they can design airliners on computer.
Likewise, my lab and our sister lab have not done a test in 16 years; because we can
simulate them on the computer.
post #316 of 481
Quote:
Originally Posted by Alimentall View Post

Did you miss where it was said that it is simply an equation? That if you 'cheat' by removing some of the problems, that it throws the equation off and you end up with a poorer result?

John,

What are you throwing out John?
Quote:


Sure, but the math is a tool. The process itself is not an equation. It's like you saying that driving to work is an equation. And that if you put anti-lock brakes on your car, that it is more dangerous because it takes away one of the problems with driving.

If you are talking about unconstrained degrees of freedom - those ARE a problem in
computational physics. If the system can "move" or "do something" and you have not
given the system a rule to follow to dictate that motion; then your simulation will just
give random numbers.
post #317 of 481
Thread Starter 
A-ha. So you're saying you *could* model speaker design as opposed to 'it's an equation'. Until it actually *is* a series of equations, a full simulation, there is nothing to get 'under determined' in the simulation that doesn't exist. Because if you actually do a real simulation, you'll find that adding advanced DSP into the simulation will cause some of it to get 'zeroed out' and the simulation suddenly become simpler, just as it is in real life.

Such a simulation would have to have all the driver measurements for every driver in existence, as well as the ability to custom order or design drivers from scratch, shape the cabinet, make it have some WAF factor for those who have wives, etc, etc. It doesn't exist, therefore, humans still do all the hard work and use tools to accomplish their goals, the most advanced tool now available being DSP crossover/impulse response correction, aka DEQX.
post #318 of 481
Quote:
Originally Posted by John Kotches View Post

Our current physics will do until the next set of even better theories surpass them. It took about 250 years to rewrite Newton with General Relativity. How long before General Relativity is rewritten (as an example).

Pardon me being slightly tongue cheek here. Physics (as you know very very well) is ever evolving

John,

Yes - at the level of the attempts to unify all the forces; then yes it is evolving. That is
a few years ago we had 4 basic forces; gravity, the strong nuclear force [ holds the
nucleus together and is responsible for the energy that nuclear reactors release ], the
weak nuclear force [ responsible for some radioactive decay ], and the Coulomb force
[ which is the force that provides the force between charged particles - like the nucleus
and the electrons; and hence is responsible for all chemistry. Any chemical reaction,
like combustion that releases energy is due to the this force ].

With the discovery of the intermediate vector boson or "W" some time back; we now
know that the weak nuclear and the Coulomb force are two manifestations of the same
force - now called the electro-weak force.

The aim is to unify all these forces; to say that all are just different manifestations of one
force. That is called Grand Unification.

However, physics on the scale of loudspeakers, and how they move, the electric forces
that drive them, how the air responds.... is all very well known physics that has been
known for over a hundred years. There has been nothing new in the understanding of
the physics that underlie loudspeakers since the mid 19-th century.
post #319 of 481
Thread Starter 
Quote:
Originally Posted by Morbius View Post

If you are talking about unconstrained degrees of freedom - those ARE a problem in
computational physics. If the system can "move" or "do something" and you have not
given the system a rule to follow to dictate that motion; then your simulation will just
give random numbers.

Yes, but speaker design still isn't an equation. The DEQX box obviously has 'rules' to generate a proper, though not necessarily perfect result. When I say the *engineer* has more degrees of freedom in solving the 'equation' (really a problem), such as the ability to use different size/shape/materials/impedances and crossover slopes for the drivers than he did before. More ways of solving the problem.

In speaker design, every choice you make essentially dictates another choice for you. Start down one path and you almost automatically end up someplace, often where you don't wan to be. B&W expresses this in their 800D white paper. Bringing DSP means that you can more easily end up where you want to be because each choice allows more 'degrees of freedom' (pathways) to achieve the solution you want and 'solve' the 'equation'.
post #320 of 481
Quote:
Originally Posted by Alimentall View Post

Could you please explain what I am BSing about? And what point I am supposed to concede? I mean, I don't mind conceding if I'm wrong. You can not say that being able to correct, for instance, time in a speaker design electrically is going to cause some bad mojo because 'mother nature' can't stand having, what was it, an 'under determined equation' (?). That's why I say a speaker design ≠ an equation. Not without being overly simplistic. DSP is not a 'cheat'. It's an advancement. It does not through nature out of balance or equations out of whack. It solves problems. Elegantly. Leaving speaker designers to focus on other things that DSP can't solve. I'm sorry if you can't follow the logic on this, but there's no need to be angry, nor to lob insults.

John,

Frantz is EXACTLY CORRECT!!! Just because YOU don't understand the science
and the mathematics - don't claim that mathematical simulation is being overly simplistic.

Just because you don't understand constraints and degrees of freedom - there are plenty
of engineers and scientists who do. You are NOT some great expert and anything outside
your understanding must be baloney.

There are people that know far more that you do about mathematics, physics, and the
design and operation of audio equipment. We have members here that have college
degrees, and even advanced degrees in the sciences here.

You don't become a fellow of the National Academy of Science and Engineering; or
even the Audio Engineering Society by being an audio salesman for 20 years.
post #321 of 481
Thread Starter 
Quote:
Originally Posted by Morbius View Post

John,

Frantz is EXACTLY CORRECT!!! Just because YOU don't understand the science
and the mathematics - don't claim that mathematical simulation is being overly simplistic.

There is no speaker design simulation!
Quote:



Just because you don't understand constraints and degrees of freedom - there are plenty
of engineers and scientists who do. You are NOT some great expert and anything outside
your understanding must be baloney.

So why do you have problems coming up with better reasoning than 'baloney'?.
Quote:



There are people that know far more that you do about mathematics, physics, and the
design and operation of audio equipment. We have members here that have college
degrees, and even advanced degrees in the sciences here.

But few know more about what's going on in the field itself. You certainly don't. You can't even answer the basic challenges I give you, such as explaining why TacT is better than DEQX.
Quote:



You don't become a fellow of the National Academy of Science and Engineering; or
even the Audio Engineering Society by being an audio salesman for 20 years.

And you don't become an expert on everything by getting a PhD in math. Want to explain your 'half-wave' theory of setting up dipoles again for us?
post #322 of 481
Quote:
Originally Posted by Alimentall View Post

Yes, but speaker design still isn't an equation. .

John.

It is not a single equation - but it IS a set of coupled differential equations.

Why don't you take a few years [ or decades ] to get a degree in engineering and / or
science with the requisite mathematics? Then maybe we can discuss this.

You can hand wave all you want - but in the final analysis; every component in an audio
system obey physical laws that can be cast as equations. I don't know why you can't
understand that!!!

How do you think we launch spacecraft that can rendezvous with the outer planets many
years later? Do we take your approach and say, "Oh, a spacecraft is not an equation" or
"Oh, a planet is not an equation". NO - the motion of both spacecraft and planets can
be described by mathematics; and by solving those equations - we can put a spacecraft
in a region of space nearby those planets.

You don't know how to do that; but just because you don't know how to do it; don't say
that others don't know how to do it or that it can't be represented that way.
post #323 of 481
Quote:
Originally Posted by Alimentall View Post

A-ha. So you're saying you *could* model speaker design as opposed to 'it's an equation'. Until it actually *is* a series of equations, a full simulation, there is nothing to get 'under determined' in the simulation that doesn't exist. Because if you actually do a real simulation, you'll find that adding advanced DSP into the simulation will cause some of it to get 'zeroed out' and the simulation suddenly become simpler, just as it is in real life.

John,

There ARE computer programs that can model this in existence. A speaker is basically
a mechanical system. There are computer programs that can EASILY handle such
mechanical systems. My lab is responsible for one of the most used and most
advanced; DYNA3D:

https://www-eng.llnl.gov/mdg/mdg_codes_dyna3d.html

DYNA3D has been used in the design of jet engines; like that pictured, and has been
used to analyze the earthquake response of the Golden Gate and Oakland-Bay bridges.
It could CERTAINLY handle a loudspeaker driver.

Now what are blathering about something getting "zeroed out" when you use DSP.
Be specific - WHAT gets "zeroed" because of DSP?

Quote:


Such a simulation would have to have all the driver measurements for every driver in existence, as well as the ability to custom order or design drivers from scratch, shape the cabinet, make it have some WAF factor for those who have wives, etc, etc. It doesn't exist, therefore, humans still do all the hard work and use tools to accomplish their goals, the most advanced tool now available being DSP crossover/impulse response correction, aka DEQX.

No - you don't have to know about every driver in existence. YOU tell the computer
what the driver in the speaker you are designing looks like - and what its properties are;
compliance of the cone suspension, for example. It computes how that driver moves.
Couple that to a wave equation, and you can calculate the soundfield.

Don't like the result? Change the design of the driver - change the compliance and
redo the calculation. Once you get something you like; THEN you can hand the specs
over to craftsmen to actually make the part.

This is how jet engines, airliners, cars....LOTS of things are designed.

Why would you "think" that you would need to know the parameters of every driver
in existence??? Why would you even consider that you needed that? You are designing
a tweeter for NHT; why would you need to know what a woofer being made in Germany
looks like?

Come on, John; put on your "thinking cap".
post #324 of 481
Thread Starter 
Quote:
Originally Posted by Morbius View Post

It is not a single equation - but it IS a set of coupled differential equations.

A very large set of possible equations that are, as yet, a figment of your imagination. Unless you'd like to link to them or take a few pages and actually write these equations.
Quote:



Why don't you take a few years [ or decades ] to get a degree in engineering and / or
science with the requisite mathematics? Then maybe we can discuss this.

Why don't you get off your high horse and and stop spouting pseudo-science?
Quote:



You can hand wave all you want - but in the final analysis; every component in an audio
system obey physical laws that can be cast as equations. I don't know why you can't
understand that!!!

I do understand that, but I also understand that DSP gives you more room to maneuver and come back with a closer to ideal result.
Quote:



How do you think we launch spacecraft that can rendezvous with the outer planets many
years later? Do we take your approach and say, "Oh, a spacecraft is not an equation" or
"Oh, a planet is not an equation". NO - the motion of both spacecraft and planets can
be described by mathematics; and by solving those equations - we can put a spacecraft
in a region of space nearby those planets.

That is a much simpler set of equations than creating one for building a speaker. Those equations and simulations actually exist, the one you keep talking about for speakers does not.
Quote:



You don't know how to do that; but just because you don't know how to do it; don't say
that others don't know how to do it or that it can't be represented that way.

I didn't say that. I'm saying the equations don't exist and you're just making stuff up to try to talk above everyone. Doesn't work with me. DSP solves problems, it doesn't create underdetermined equations.
post #325 of 481
Quote:
Originally Posted by Alimentall View Post

There is no speaker design simulation!

John,

I could setup a calculation of a loudspeaker driver the same way that jet engine
simulation was set up. Do you honestly think a loudspeaker driver is more complex
than a jet engine?

Quote:


But few know more about what's going on in the field itself. You certainly don't. You can't even answer the basic challenges I give you, such as explaining why TacT is better than DEQX.

Sure - I already gave you that answer. The mathematics implemented in the TacT is
WAY WAY better than what the DEQX has. Additionally, as I stated, TacT has mastered
Dynamic Room Correction [DRC] and Audyssey has followed suit. DEQX is mssing in
action on that one.
Quote:


And you don't become an expert on everything by getting a PhD in math. Want to explain your 'half-wave' theory of setting up dipoles again for us?

That theory is CORRECT - the fact that YOU don't understand it not withstanding.

John, you've proved time and again that you are absolutely clueless on the mathematics
of the wave equation - so it's no wonder you don't understand that theory.

Get an education John - and quit telling people who are more educated than you are that
they don't know what they are talking about.

Over and out.
post #326 of 481
Thread Starter 
Quote:
Originally Posted by Morbius View Post

There ARE computer programs that can model this in existence. A speaker is basically
a mechanical system. There are computer programs that can EASILY handle such
mechanical systems. My lab is responsible for one of the most used and most
advanced; DYNA3D:

Sure, how many are programmed to do it? None.
Quote:



Now what are blathering about something getting "zeroed out" when you use DSP.
Be specific - WHAT gets "zeroed" because of DSP?

Time/phase, for one.
Quote:



No - you don't have to know about every driver in existence. YOU tell the computer
what the driver in the speaker you are designing looks like - and what its properties are;
compliance of the cone suspension, for example. It computes how that driver moves.
Couple that to a wave equation, and you can calculate the soundfield.

So, in the end, you still have a human using a tool, rather than a 'set of equations' determining the result.
Quote:



Come on, John; put on your "thinking cap".

Sure. You put on your 'logic cap' first. And your 'reasoning slippers'. A 'reality cape' might be helpful.
post #327 of 481
Thread Starter 
Quote:
Originally Posted by Morbius View Post

John,

I could setup a calculation of a loudspeaker driver the same way that jet engine
simulation was set up. Do you honestly think a loudspeaker driver is more complex
than a jet engine?

A jet engine may be more complex than a speaker driver, but not as complex as a speaker system. Jet engines are pretty simple creatures.
Quote:




Sure - I already gave you that answer. The mathematics implemented in the TacT is
WAY WAY better than what the DEQX has.

Because you say so? Have go gone through all the lines of code in both? Specifics, please, not just 'hope'.
Quote:



Additionally, as I stated, TacT has mastered
Dynamic Room Correction [DRC] and Audyssey has followed suit. DEQX is mssing in
action on that one.

DEQX is not room correction! It is a speaker correction device with parametric EQ as well. The computer in my car's engine doesn't play the violin.
Quote:



That theory is CORRECT - the fact that YOU don't understand it not withstanding.

Oh, I understand what you think you're doing, but you are, again, being incredibly simplistic, looking to math rather than psychocoustics.
Quote:



John, you've proved time and again that you are absolutely clueless on the mathematics
of the wave equation - so it's no wonder you don't understand that theory.

Get an education John - and quit telling people who are more educated than you are that
they don't know what they are talking about.

Over and out.

For a bright guy, you keep trying to show how dumb you are. Lots of bluster, but no actual facts.
post #328 of 481
It'd be nice to see the personal attacks/diatribes end. Perhaps BOTH sides could simply ignore the baiting and let it (finally) end.

I personally do not like the dynamic loudness in the Audyssey or the Tact gear. Maybe my ears have gotten used to how my system sounds at various volume levels.

Speaker/driver correction and even more so PEQ for room anomalies below 200hz, that's another matter.
post #329 of 481
Thread Starter 
BTW, I found this 4-way DSP crossover w/up to 48dB/octave slopes -

http://www.ashly.com/protea424c.html

I also saw that there is/was a company working on a 5-way DEQX solution for the Magico Ulimate speaker. Not that it would need it, right
post #330 of 481
Thread Starter 
Quote:
Originally Posted by Brucemck2 View Post

I personally do not like the dynamic loudness in the Audyssey or the Tact gear. Maybe my ears have gotten used to how my system sounds at various volume levels.

I was speaking with NAD about this and they said that Audyssey Dynamic EQ/Loudness comes in too aggressively and quickly, so they had it toned way down to be much more subtle. I think with mass market technologies, people like to hear what they are buying, the same reason subs and rears always seem to be 5-10dB too loud when we go to peoples' homes.
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Ultra Hi-End HT Gear ($20,000+)
This thread is locked  
AVS › AVS Forum › Display Devices › Ultra Hi-End HT Gear ($20,000+) › Digital Active Speaker Thread.....